Amino Acid Analysis Protocols (Methods in Molecular Biology)

Methods in Molecular Biology TM VOLUME 159 Amino Acid Analysis Protocols Edited by Catherine Cooper Nicolle Packer K...

Author: Catherine Cooper

10 downloads 531 Views 1MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Methods in Molecular Biology

TM

VOLUME 159

Amino Acid Analysis Protocols Edited by

Catherine Cooper Nicolle Packer Keith Williams

HUMANA PRESS

Amino Acid Analysis

1

1 Amino Acid Analysis An Overview Margaret I. Tyler 1. Importance and Utility Amino acids are found either in the free state or as linear chains in peptides and proteins. There are 20 commonly occurring amino acids in proteins, which are shown in Table 1. Amino acid analysis has an important role in the study of the composition of proteins, foods, and feedstuffs. Free amino acids are also determined in biological material, such as plasma and urine, and in fruit juice and wine. When it is performed on a pure protein, amino acid analysis is capable of identifying the protein (2,3, and Chapter 8 in this volume), and the analysis is also used as a prerequisite for Edman degradation and mass spectrometry and to determine the most suitable enzymatic or chemical digestion method for further study of the protein. It is also a useful method for quantitating the amount of protein in a sample (see Chapter 2 in this volume) and can give more accurate results than colorimetric methods. 2. Historical View The earliest experiments on the acid hydrolysis of proteins were performed by Braconnot in 1820, in which concentrated sulphuric acid was used to hydrolyze gelatin, wool, and muscle fibers (4). Various reagents for performing protein hydrolysis were tried over the next 100 years, with 6 M HCl becoming the most widely accepted reagent. In 1972, Moore and Stein (5) were awarded the Nobel Prize for developing an automated instrument for separation of amino acids on an ion-exchange resin and quantitation of them using ninhydrin. More recently, high-performance liquid chromatographs (HPLCs) have been configured for amino acid analysis. Some methods use postcolumn derivFrom: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

1

2

Tyler

Table 1 Common Amino Acids

3 letter

1 letter

Essential for humans (1)

Asp Glu

D E

No No

Ala Asn Cys Gln Gly Ile Leu Met Phe Ser Thr Trp Tyr Val

A N C Q G I L M F S T W Y V

No No No No No Yes Yes Yes Yes No Yes Yes No Yes

Arg His Lys

R H K

Yes Yes Yes

Pro

P

No

Symbol Name Acidic amino acids Aspartic acid Glutamic acid Neutral amino acids Alanine Asparagine Cysteine Glutamine Glycine Isoleucine Leucine Methionine Phenylalanine Serine Threonine Tryptophan Tyrosine Valine Basic amino acids Arginine Histidine Lysine Imino acid Proline

atization in which the amino acids are separated on an ion-exchange column followed by derivatization with ninhydrin (6, and Chapter 2 in this volume), fluorescamine (7), or o-phthalaldehyde (8). Another approach has been to derivatize amino acids prior to separation on a reversed-phase HPLC column. Examples of this technique are dansyl (9), phenylisothiocyanate (PITC) (10, and Chapters 12 and 13 of this volume), 9-fluorenylmethyl chloroformate (Fmoc) (11), and 6-aminoquinolyl-N-hydroxysuccinimyl carbamate (AQC) (12, and Chapters 4 and 8 in this volume). 3. Sensitivity Amino acid analysis can be performed accurately at the fmol level by methods employing fluorescence detection, whereas for derivatives detected by

Amino Acid Analysis

3

Table 2 Comparison of Different Derivatization Chemistries for Amino Acid Analysis Derivatization typea Detection modeb Sensitivity Chromatographyc

Ninhydrin

OPA

OPA

PITC

Fmoc

AQC

postc c pmol i.e.

postc f fmol i.e.

prec f fmol r.p.

prec UV pmol r.p.

prec f fmol r.p.

prec f fmol r.p

OPA, orthophthalaldehyde; PITC, phenylisothiocyanate; Fmoc, 9-fluorenylmethyl chloroformate; AQC, 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate. apostc,

postcolumn; prec, precolumn. colorimetry; f, fluorescence; UV, ultraviolet. ci.e, ion exchange; r.p., reversed-phase HPLC. bc,

ultraviolet (UV) light, the analysis is at the pmol level. Table 2 gives a comparison of the various derivatization chemistries and their sensitivities. Annual studies comparing the various methods have been carried out by the Association of Biomolecular Resource Facilities (ABRF) (13,14). Strydom and Cohen (15) have compared AQC and PITC chemistries and found AQC derivatives to be more stable.

4. Difficult Amino Acids 4.1. Tryptophan Tryptophan is destroyed in acid hydrolysis. Alkaline hydrolysis with NaOH, Ba (OH)2, or LiOH have been used particularly in the hydrolysis of food and feedstuffs (16,17). However, acid hydrolysis is still needed to determine the other amino acids. There have been a number of methods published for the determination of tryptophan that use the standard 6 M HCl hydrolysis in the presence of additives, some of which include thioglycolic acid (18), beta-mercapto ethanol (19), and mercaptoethanesulfonic acid (20).

4.2. Cysteine and Cystine Cysteine and cystine are unstable during acid hydrolysis, particularly in the presence of carbohydrate. The total content of cysteine and cystine can be determined by oxidizing the protein with performic acid, which converts both forms to cysteic acid and methionine to methionine sulphone. The protein is then hydrolyzed with 6 M HCl (17). Disulphide compounds such as dithiopropionic acid and dithiobutyric acid have been proposed as protecting agents for cysteine and cystine during acid

4

Tyler

hydrolysis (21). The use of dithiodiglycolic acid during the acid hydrolysis step, followed by phenylisothiocyanate derivatization, allows cysteine and cystine plus all common hydrolysate amino acids (excluding tryptophan) to be determined (22). Reduction of disulphide bridges, followed by alkylation of cysteine with iodoacetic acid or 4-vinylpyridine is also used to determine the cysteine-pluscystine content of proteins (23). Alkylation with acrylamide produces cysteineS-propionamide, which is converted to cysteine-S-propionic acid during acid hydrolysis (24).

4.3. Asparagine and Glutamine Asparagine and glutamine are amide derivatives of aspartic acid and glutamic acid, respectively. During acid hydrolysis, which cleaves amide bonds, asparagine is converted to aspartic acid and glutamine to glutamic acid. Thus, the amount determined for aspartic acid represents the total of aspartic acid and asparagine and similarly for glutamic acid and glutamine. 4.4. D amino Acids The D-amino acid content of a protein or peptide can be determined by employing a short partial acid hydrolysis, followed by an enzymatic hydrolysis with pronase, and then with leucine aminopeptidase and peptidyl-D-amino acid hydrolase (25). 5. Modified Amino Acids Phosphorylated amino acids are able to be analyzed using a variety of different chemistries (26), but the ABRF 1993 study found that precolumn methods were more successful (27). Phosphoserine (28,29), phosphothreonine (29), and phosphotyrosine (29) have varying stabilities. Highest recoveries for phosphoserine and phosphotyrosine are produced with hydrolysis time of 60 min or less at 110ºC, whereas for phosphothreonine, a hydrolysis time of 2 h gave better results (26). Chapter 14 in this volume covers the analysis of phosphoamino acids more extensively. There are many other rarer amino acids and derivatives that can be analyzed. These include hydroxyproline (17,30, and Chapter 16 in this volume) and hydoxylysine (17,30, and Chapter 2 of this volume), found in collagen. Taurine has dietary importance and can be readily determined in infant formulas, pet food, plasma, urine, and tissue extracts (17, and Chapter 10 of this volume). Posttranslational modifications, including glycosylated amino acids (31, and Chapters 2 and 7 of this volume) and glycated amino acids (32,33), are important in studying protein function. Chapter 18 of this volume describes the application of mass spectrometry to the analysis of glycated amino acids.

Amino Acid Analysis

5

6. Limitations—Contaminants and Precautions The accuracy of amino acid analysis is very dependent on the integrity of the sample. Cleanliness of all surfaces the sample contacts is essential, as is the purity of all reagents used. Traces of salts, metals, or detergents can effect the accuracy of results. The hydrolysis step is particularly important, as was demonstrated in the ABRF 1994 AAA collaborative study (34). Many laboratories now satisfactorily perform a 1-h hydrolysis in 6 M HCl at 150°C under vacuum. The traditional method uses 6 M HCl for 20–24 h at 110°C under vacuum. Losses of serine, threonine, and to a lesser extent, tyrosine may occur under these conditions. During acid hydrolysis, some amide bonds between aliphatic amino acids are more difficult to cleave. The Ala–Ala, Ile–Ile, Val–Val, Val–Ile, Ile–Val, and Ala–Val linkages are resistant to hydrolysis and may need a longer hydrolysis time of 48 or 72 h at 110ºC (35). References 1. Encyclopaedia of Food Science Food Technology and Nutrition, vol. 1 (Macrae, R., Robinson, R. K., and Sadler, M. J., eds.), Academic, London, p. 149. 2. Hobohm, U., Houthaeve, T., and Sander, C. (1994) Amino acid analysis and protein database compositional search as a rapid and inexpensive method to identify proteins. Anal. Biochem. 222, 202–209. 3. Schegg, K. M., Denslow, N. D., Andersen, T. T., Bao, Y. A., Cohen, S. A., Mahrenholz, A. M., and Mann, K. (1997) Quantitation and identification of proteins by amino acid analysis: ABRF-96 collaborative trial, in Techniques in Protein Chemistry VIII (Marshak, D., ed.), Academic, San Diego, CA, pp. 207–216. 4. Braconnot, H. (1820) Ann. Chim. Phys. 13, 113. 5. Moore, S. and Stein, W. H. (1963) Chromatographic determination of amino acids by the use of automatic recording equipment, in Methods in Enzymology, vol. 6 (Colowick, S. P. and Kaplan, N. O., eds.), Academic, New York, pp. 819–831. 6. Samejima, K., Dairman, W., and Udenfriend, S. (1971) Condensation of ninhydrin with aldehydes and primary amines to yield highly fluorescent ternary products. 1. Studies on the mechanism of the reaction and some characteristics of the condensation product. Anal. Biochem. 42, 222–236. 7. Stein, S., Bohlen, P., Stone, J., Dairman, W., and Udenfriend, S. (1973) Amino acid analysis with fluorescamine at the picomole level. Arch. Biochem. Biophys. 155, 202–212. 8. Roth, M. (1971) Fluorescence reaction for amino acids. Anal. Chem. 43, 880–882. 9. Tapuhi, Y., Schmidt, D. E., Lindner, W., and Karger, B. L. (1981) Dansylation of amino acids for high-performance liquid chromatography analysis. Anal. Biochem. 115, 123–129. 10. Bidlingmeyer, B. A., Cohen, S. A., and Tarvin, T. (1984) Rapid analysis of amino acids using pre-column derivatisation. J. Chromatog. 336, 93–104.

6

Tyler

11. Haynes, P. A., Sheumack, D., Kibby, J., and Redmond, J. W. (1991) Amino acid analysis using derivatisation with 9-fluorenylmethyl chloroformate and reversedphase high performance liquid chromatography. J. Chromatog. 540, 177–185. 12. Strydom, D. J. and Cohen, S. A. (1993) in Techniques in Protein Chemistry IV (Angeletti, R. H., ed.), Academic, San Diego, CA, pp. 299–307. 13. Marenholz, A. M., Denslow, N. D., Andersen, T. T., Schegg, K. M., Mann, K., Cohen, S. A., et al. (1996) Amino acid analysis — recovery from PVDF membranes: ABRF-95AAA collaborative trial, in Techniques in Protein Chemistry VII (Marsak, D. R., ed.), Academic, San Diego, CA, pp. 323–330. 14. Tarr, G. E., Paxton, R. J., Pan, Y. C.-E, Ericsson, L. H., and Crabb, J. W. (1991) Amino acid analysis 1990: the third collaborative study from the association of biomolecular resource facilities (ABRF) in Techniques in Protein Chemistry II (Villafranca, J. J., ed.), Academic, San Diego, CA, pp. 139–150. 15. Strydom, D. J. and Cohen, S. A. (1994) Comparison of amino acid analyses by phenylisothiocyanate and 6-aminoquinolyl-N-hydroxysuccinimyl carbamate precolumn derivatisation. Anal. Biochem. 222, 19–28. 16. Delhaye, S. and Landry, J. (1986) High-performance liquid chromatography and ultraviolet spectrophotometry for quantitation of tryptophan in barytic hydrolysates. Anal. Biochem. 159, 175–178. 17. Cohen, S. A., Meys, M., and Tarvin, T. L. (1988) The PicoTag Method. A Manual of Advanced Techniques for Amino Acid Analysis. Waters Chromatography Division, Millipore Corp., Milford, MA. 18. Yokote, Y., Murayama, A., and Akahane, K. (1985) Recovery of tryptophan from 25-minute acid hydrolysates of protein. Anal. Biochem. 152, 245–249. 19. Ng, L. T., Pascaud, A., and Pascaud, M. (1987) Hydrochloric acid hydrolysis of proteins and determination of tryptophan by reversed-phase high-performance liquid chromatography. Anal. Biochem. 167, 47–52. 20. Yamada, H, Moriya, H., and Tsugita, A. (1991) Development of an acid hydrolysis method with high recoveries of tryptophan and cysteine for microquantities of protein. Anal. Biochem. 198, 1–5. 21. Barkholt, V. and Jensen, A. L. (1989) Amino acid analysis: determination of cysteine plus half-cystine in proteins after hydrochloric acid hydrolysis with a disulphide compound as additive. Anal. Biochem. 177, 318–322. 22. Hoogerheide, J. G. and Campbell, C. M. (1992) Determination of cysteine plus half-cystine in protein and peptide hydrolysates: use of dithiodiglycolic acid and phenylisothiocyanate derivatisation. Anal. Biochem. 201, 146–151. 23. Inglis, A. S. (1983) Single hydrolysis method for all amino acids, including cysteine and tryptophan, in Methods in Enzymology, vol. 91. Academic, San Diego, CA, pp. 26–36. 24. Yan, J. X., Kett, W. C., Herbert, B. R., Gooley, A. A., Packer, N. H., and Williams, K. L. (1998) Identification and quantitation of cysteine in proteins separated by gel electrophoresis. J. Chromatog. 813, 187–200.

Amino Acid Analysis

7

25. D’Aniello, A., Petrucelli, L., Gardner, C., and Fisher, G. (1993) Improved method for hydrolysing proteins and peptides without introducing racemization and for determining their true D-amino acid content. Anal. Biochem. 213, 290–295. 26. Yan, J. X., Packer, N. H., Gooley, A. A., and Williams, K. L. (1998) Protein phosphorylation: technologies for the identification of phosphoamino acids. J. Chromatog. A. 808, 23–41. 27. Yüksel, K. Ü., Andersen, T. T., Apostol, I., Fox, J. W., Crabb, J. W., Paxton, R. J., and Strydom, D. J. (1994) Amino acid analysis of phospho-peptides: ABRF-93AAA, in Techniques in Protein Chemistry V (Crabb, J. W., ed.), Academic, San Diego, CA, pp. 231–240. 28. Meyer, H. E., Swiderek, K., Hoffmann-Posorske, E., Korte, H., and Heilmeyer, L. M., Jr. (1987) Quantitative determination of phosphoserine by high-performance liquid chromatography as the phenylthiocarbamyl-S-ethylcysteine. Application to picomolar amounts of peptides and proteins. J. Chromatog. 397, 113–121. 29. Ringer, D. P. (1991) Separation of phosphotyrosine, phosphoserine and phosphothreonine by high-performance liquid chromatography, in Methods in Enzymology, vol. 201. Academic, San Diego, CA, pp. 3–10. 30. Waters AccQ.Tag Amino Acid Analysis System Operators Manual (1993). Millipore Corp., Melford, MA. 31. Packer, N. H., Lawson, M. A., Jardine, D. R., Sanchez, J. C., and Gooley, A. A. (1998) Analyzing glycoproteins separated by two-dimensional gel electrophoresis. Electrophoresis 19, 981–988. 32. Walton, D. J. and McPherson, J. D. (1987) Analysis of glycated amino acids by high-performance liquid chromatography of phenylthiocarbamyl derivatives. Anal. Biochem. 164, 547–553. 33. Cayot, P. and Tainturier, G. (1997) The quantitation of protein amino groups by the trinitrobenzenesulfonic acid method: a reexamination. Anal. Biochem. 249, 184– 200. 34. Yüksel, K. Ü., Andersen, T. T., Apostol, I., Fox, J. W., Crabb, J. W., Paxton, R. J., and Strydom, D. J. (1994) The hydrolysis process and the quality of amino acid analysis: ABRF-94AAA collaborative trial, in Techniques in Protein Chemistry VI (Crabb, J. W., ed.), Academic, San Diego, CA, pp. 185–192. 35. Ozols, J. (1990) Amino acid analysis, in Methods in Enzymology, vol. 182. Academic, San Diego, CA, pp. 587–601.

Role of AAA in a Biotechnology Laboratory

9

2 Amino Acid Analysis, Using Postcolumn Ninhydrin Detection, in a Biotechnology Laboratory Frank D. Macchi, Felicity J. Shen, Rodney G. Keck, and Reed J. Harris 1. Introduction Although lacking the speed and sensitivity of more widely heralded techniques such as mass spectrometry, amino acid analysis remains an indispensable tool in a complete biotechnology laboratory responsible for the analysis of protein pharmaceuticals. Moore and Stein developed the first automated amino acid analyzer, combining cation–exchange chromatographic separation of amino acids with postcolumn ninhydrin detection (1). Commercial instruments based on this design were introduced in the early 1960s, though many manufacturers have abandoned this technology in favor of precolumn amino acid derivatization with separations based on reversed-phase chromatography (2–4) (see Note 1). In our product development role, we still rely on amino acid analysis to generate key quantitative and qualitative data. Amino acid analysis after acid hydrolysis remains the best method for absolute protein/peptide quantitation, limited in accuracy and precision only by sample handling. We produced an Excel macro to process these data; the macro transfers and converts the amino acid molar quantities into useful values such as composition (residues per mol) and concentration. In addition, we employ several specialized amino acid analysis applications to monitor structural aspects of some of our recombinant products. De novo biosynthesis of leucine in bacteria will lead to a minor amount of norleucine (Nle) production (5), particularly if recombinant proteins are produced in fermentations that have been depleted of leucine (6). The side-chain of Nle (-CH2-CH2-CH2-CH3) is similar enough to methionine (-CH2-CH2-S-CH3) that some of the tRNAMet will be acylated by Nle, leading to incorporation of Nle at From: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

9

10

Macchi et al.

Met positions (6,7). When this occurs, Nle may be incorporated at a low level at every Met position, and amino acid analysis is often the only method able to detect this substitution. Hydroxylysine (Hyl) is a common modification of lysine residues found at -Lys-Gly- positions in collagens and collagen-like domains of modular proteins (8). This modification is also found at certain solvent-accessible -Lys-Glysites in noncollagenous proteins, usually at substoichiometric levels (9). Amino acid analysis is a useful screening technique for the identification of Hyl-containing recombinant proteins produced by mammalian cells. The analysis of recombinant proteins using carboxypeptidases may still be required to assign the C-terminus when the polypeptide chain is extensively modified, thus ruling out making a C-terminal assignment based solely on mass and N-terminal analyses, or in cases where the C-terminal peptide cannot be assigned in a peptide map. When carboxypeptidase analyses are needed, a modified amino acid analysis program is needed to resolve Gln and Asn (which are not found in acid hydrolysates) from other amino acids. Assignment of Asn-linked glycosylation sites is greatly facilitated by prior knowledge of the -Asn-Xaa-Thr/Ser/Cys- consensus sequence sites (10), and specific endoglycosidases, such as peptide:N-glycosidase F can be employed to quantitatively release all known types of Asn-linked oligosaccharides (11). O-linked sites are harder to assign, as these are found in less-stringent sequence motifs (12–14), and there is no universal endoglycosidase for O-glycans except for endo-α-N-acetylgalactosaminidase, which can only release the disaccharide Gal(β1→3)GalNAc. In addition, O-glycosylation is often substoichiometric. In mammalian cell products, at least two N-acetylglucosamine (GlcNAc) residues are found in Asn-linked oligosaccharides, whereas N-acetylgalactosamine (GalNAc) is found at the reducing terminus of the most common (mucin-type) O-linked oligosaccharides. A cation–exchange-based amino acid analyzer can easily be modified for the analysis of the amino sugars glucosamine (GlcNH2) and galactosamine (GalNH2) from acid hydrolysis of GlcNAc and GalNAc, respectively, allowing confirmation of the presence of most oligosaccharide types. In glycoproteins, HPLC fractions from peptide digests can be screened using amino sugar analysis to identify glycopeptides for further analysis. Regulated biotechnology products are usually tested for identity using HPLC maps after peptide digestion (15,16). A key aspect of the digestion step for most proteins is obtaining complete reduction of all disulfide bonds, followed by complete alkylation of cysteines without the introduction of artifacts (e.g., methionine S-alkylation) (17). Amino acid analysis can be used to monitor cysteine alkylation levels for reduced proteins, such as are obtained after alkylation with iodoacetic acid, iodoacetamide or 4-vinylpyridine.


11

2. Materials

2.1. Equipment 1. 1-mL hydrolysis ampoules (Bellco, Vineland, NJ; part number 4019-00001) (see Note 2). 2. Savant SpeedVac. 3. Oxygen/methane flame. 4. Glass knife (Bethlehem Apparatus Co., Hellertown, PA). 5. 1/4" ID × 5/8" OD Tygon tubing. 6. Model 6300 analyzers (Beckman Instruments, now Beckman Coulter, Fullerton CA). The sum of the 440 nm and 570 nm absorbances is converted to digital format using a PE Model 900 A/D converter, and the data are collected by a PE Turbochrom Model 4.1 data system (see Notes 3–5). 7. Lithium-exchange column (Beckman part number 338075, 4.6 × 200 mm).

2.2. Reagents and Solutions 1. Constant boiling (6 N) HCl ampoules are obtained from Pierce (Rockland, IL) (see Note 6). 2. Mobile phase buffers purchased from Beckman Instruments include sodium citrate buffers Na-D, Na-E, Na-F, Na-R, and Na-S; lithium citrate buffers include Li-A, Li-B, Li-C, Li-D, Li-R, and Li-S. 3. Ninhydrin kits (Nin-Rx) are also purchased from Beckman; these must be mixed thoroughly before use (usually 2 h at room temperature), and care must be taken to avoid skin discoloration because of contact with ninhydrin-containing materials. 4. Dialysis may be used to desalt samples into dilute acetic acid prepared from deionized water (Milli-Q, Millipore) and Mallinckrodt U.S.P. grade glacial acetic acid. 5. Amino acid standards: are diluted from the stock Beckman standard (part number 338088) with Na-S buffer to final concentration of 40 nmol/mL or 20 nmol/ mL (see Note 7). 6. 2 N glacial acetic acid.

3. Methods

3.1. Sample Preparation Proteins should be desalted to obtain optimal compositional data. Dialysis against 0.1% acetic acid removes salts while keeping proteins in solution, but quantitative data will often require direct hydrolysis (i.e., without dilution or sample losses introduced during dialysis). When proteins must be analyzed without desalting, neutral buffers such as 50 mM Tris can be used without compromising the results. Excipients to avoid include urea (which generates abundant ammonia during hydrolysis), sugars (which caramelize during hydrolysis), and detergents such as the polysorbate and Triton types that can

12

Macchi et al.

Table 1 Standard Amino Acid Analysis Time (min) 0.0 8.5 24.5 41.0 78.0 79.0 80.0 82.5 84.0 97.0

Event

Conditions

Sample injection Start temp. gradient Buffer change Buffer change Reagent pump Buffer change Buffer change Temperature change Reagent pump Recyle (start next run)

Na-E buffer, 48°C 48°C to 60°C in 8 min Na-E to Na-F Na-F to Na-D Ninhydrin to water Na-D to Na-R Na-R to Na-E 60°C to 48°C Water to ninhydrin

Buffer pump: 16 mL/h. Reagent pump: 8 mL/h.

damage cation–exchange columns. Samples in enzyme-linked immunosorbent assays (ELISA)-type buffers should be avoided as they typically contain albumin or gelatins, whose amino acids cannot be distinguished after hydrolysis from the protein of interest. Peptides generally can be desalted by reverse phase (RP)-HPLC using volatile solvents such as 0.1% TFA in water/ acetonitrile. 1. Place samples in hydrolysis ampoules (see Note 8), then dry under vacuum using a Savant SpeedVac. 2. Place approx 100 µL of 6 N HCl in the lower part of the ampoule (see Note 9). Freeze in a dry ice/ethanol bath, attached to a vacuum system via 1/4" ID × 5/ 8" OD Tygon tubing, then slowly thaw and evacuate to < 150 mtorr. 3. Use an oxygen/methane flame to seal the neck of the tube at the constriction. 4. Place the sealed ampoules in a 110°C oven for 24 h (see Note 10), then allow to cool before opening after scoring them with a glass knife. 5. Remove the acid by vacuum centrifugation, again using a Savant system, with a NaOH trap inserted between the centrifuge and cold trap. 6. After hydrolysis and acid removal, samples that contain 0.5–10 µg of protein, or 0.1–1 nmol of peptide fractions should be reconstituted with 60–200 µL of Na-S sample buffer (see Note 11).

3.2. Protein/Peptide Quantitation 1. Subject triplicate samples containing 0.5–10 µg of protein or 0.1–1 nmol of peptide to 24-h hydrolysis in vacuo as aforementioned. 2. Follow the standard operating conditions given in Table 1 (see Note 12). A standard chromatogram containing 2 nmol of each component is shown in Fig. 1.


13

Fig. 1. Analysis of a standard amino acid mixture. The standard contains 2 nmol of each component except for NH3. Operating parameters are given in Table 1. 3. Peak area data from the Turbochrom system are converted to nmol values by external standard calibration; internal standards are not necessary if a reliable autosampler is used. 4. The amino acid nmol values are also automatically converted to .tx0 files that can be imported into a custom Microsoft Excel program called the AAA MACRO (Table 2) for analysis using a PC-based computer. 5. The first step in running the AAA MACRO is to open a template, such as the example “protein.xls” given in Table 3. The residues per mol and molecular mass calculations must be modified and saved for each different protein/peptide; Asn and Asp are reported as Asx, whereas Gln, Glu and pyroglutamate are reported as Glx. 6. The macro asks for some background information (e.g., requestor’s name, sample name, number of replicates), sample prep information (e.g., volumes of hydrolysate loaded vs reconstitution volume, original sample volume), then processes the data, providing a single-page report showing calculated compositions and concentration, as shown in Fig. 2 (see Note 13–15).

3.3. Norleucine Incorporation 1. Detection of trace Nle levels in Escherichia coli-derived proteins require 24-h hydrolysis of 25–100 µg of protein (see Note 16).

14

Table 2 Amino Acid Analysis Data Conversion Macro Commands

14

Asks for Requestor's Name Halts macro if Cancel button is clicked Returns name to data worksheet cell B4 Asks for Requestor's Extension Halts macro if Cancel button is clicked Returns extension to data worksheet cell B5 Asks for protein to be analyzed Halts macro if Cancel button is clicked Returns protein's name to data worksheet cell E4 Asks for Requestor's MailStop Halts macro if Cancel button is clicked Returns Requestor's MailStop to data worksheet cell E5 Asks for MW Halts macro if Cancel button is clicked Returns MW to data worksheet cell B28 Asks for amount sample put in ampoule

Asks for reconstitution volume of sample Halts macro if Cancel button is clicked

Macchi et al.

Macro 4(a) =ACTIVATE("MACRO4A.XLM") =HIDE() =\TC4\DATA =SELECT(!B4) =INPUT("Requestor's Name?",2,"Name","" =IF(A7=FALSE,HALT()) =FORMULA(A7) =SELECT(!B5) =INPUT("Requestor's Extension?",1,"Telephone extension","") =IF(A11=FALSE,HALT()) =FORMULA(A11) =SELECT(!E4) =INPUT("Sample to be analyzed?",2,"Sample name", "") =IF(A15=FALSE,HALT()) =FORMULA(A15) =SELECT(!E5) =INPUT("Requestor's Mail Stop?",1,"Mail Stop","") =IF(A19=FALSE,HALT()) =FORMULA(A19) =SELECT(!B28) =INPUT("Molecular Mass of protein to be analyzed?",1, "Molecular Mass (g/mole)","") =IF(A23=FALSE,HALT()) =FORMULA(A23) =SELECT(!F29) =INPUT("µL in Ampoule?",1,"Ampoule volume (µL)","") =IF(A27=FALSE,HALT()) =FORMULA(A27) =SELECT(!F30) =INPUT("Sample reconstitution volume?",1,"Reconstituted volume (µL)","") =IF(A31=FALSE,HALT()) =FORMULA(A31) =SELECT(!F31)

Selects cell B1 on worksheet Asks how many replicate samples will be processed Halts macro if Cancel button is clicked Places users sample number in cell B2 Resets counter1 Selects active cell to be B6 Selects disk in drive as AAA directory Opens AAA disk Top of loop and Select data file from listed files Select data file to open Copies AAA nanomole data Closes data file

Pastes nanomole values into worksheet Adds value of 1 to the counter1 Checks to see what value = counter1 if >=1 then loops up to the top of the loop, if 0 then proceeds downward Asks for name of first data file chosen

Asks for name of second data file chosen

Asks for name of third data file chosen Asks if you want to save data Halts macro if Cancel button is clicked

Halts the macro

15

=SELECT(!C40) =INPUT("What is the name of your first replicate?",2, "Name of 1st replicate","") =IF(A56=FALSE,HALT()) =FORMULA(A56) =SELECT(!D40) =INPUT("What is the name of your second replicate?",2, "Name of 2nd replicate","") =IF(A60=FALSE,HALT()) =FORMULA(A60) =SELECT(!E40) =INPUT("What is the name of your third replicate?",2, "Name of 3rd replicate","") =IF(A64=FALSE,HALT()) =FORMULA(A64) =SAVE.AS?(,1) =IF(A67=FALSE,HALT()) =ACTIVATE("MACRO4A.XLM") =UNHIDE() =ACTIVATE.NEXT() =RETURN()

Asks if samples were diluted prior to analysis Halts macro if Cancel button is clicked


15

=INPUT("Dilution Factor?",1,"Dilution factor","" =IF(A35=FALSE,HALT()) =FORMULA(A35) =SELECT(!B1) =INPUT("How many replicates will you be analyzing today?", 1,"Number of replicates","") =IF(A39=FALSE,HALT()) =FORMULA(A39) =SET.NAME("counter1",1) =SELECT(!B41) =DIRECTORY("\TC41\data") =FILES("*.*") =OPEN?("*.*",0,FALSE,2) =SELECT("R38C6:R57C6") =COPY() =CLOSE() =ACTIVATE.NEXT() =SELECT("RC[1]") =PASTE() =SET.NAME("counter1",counter1+1) =IF(counter110 µL of isoBCF. 4. All amino, imino, imidazole, phenolic hydroxyl, and sulfhydryl groups are substituted with isoBOC groups, whereas alcoholic hydroxyl groups, guanidino group, and indole ring nitrogen, as well as amide groups are not.

Capillary Gas Chromatography of Amino Acids

109

5. The reaction mixture was washed with diethyl ether under alkaline condition in order to remove the excess reagent. This procedure also serves to exclude amines and phenols both of which often coexist with amino acids in biological samples, as they are derivatized to the corresponding N- or O-isoBOC derivatives, which are soluble in organic solvents under the same conditions as aforementioned (45–47). 6. The resulting N(O,S)-isoBOC amino acids in aqueous layer could be quantitatively and selectively extracted into diethyl ether after acidification to pH 1.0– 2.0. In these extractions, peroxide-free diethyl ether should be employed as an extraction solvent to obviate decomposition of sulfur amino acids, and great care should be taken for this solvent when small amounts of these amino acids are to be analyzed by NPD-GC. Furthermore, the ether layers should be collected, taking care to avoid aqueous droplets. It was not necessary to complete draw the ether layer in each extraction. 7. The subsequent methylation of the ethereal extracts could be successfully carried out by bubbling diazomethane. This reaction should be performed in a well-ventilated hood because diazomethane is explosive and toxic. 8. The derivative preparation could be performed within 10 min and the N(O,S)isoBOC methyl ester derivatives of protein and nonprotein amino acids were very stable under normal laboratory conditions, and no decomposition was observed during GC analysis. 9. Arg requires conversion into Orn with arginase prior to derivatization, because the guanidino group of Arg cannot be derivatized by this method and this amino acid therefore cannot be extracted into diethyl ether because of its polarity. 10. In this reaction, the conversion yield of Arg into Orn was above 93% at each level of Arg from 0.5–10 µg. 11. In a preliminary test for several megabore capillary columns, DB-17 and DB1701 gave best separation for protein amino acids, although some pairs of amino acids were partially overlapped, e.g., Thr-Ser in DB-17 column, and Leu-Ile and Glu-Met in DB-1701 column (37). Therefore, we tested DB-17 narrow-bore columns at length of 5–20 m to obtain optimum analytical conditions with respect to the separation of amino acids and analysis time. Although the resolution of amino acids increased according to increasing length of column, the analysis time became long. Among the columns tested, 15-m column of DB-17 was proved to be best column for these purposes. As shown in Fig. 4, 22 protein amino acids could be completely resolved as single and symmetrical peaks within 9 min on a single capillary column (DB-17, 15 m × 0.25-mm id, 0.25-µm film thickness) (37). However, the separation of protein and nonprotein amino acids were incomplete on this column. Furthermore, high boiling point derivatives were eluted slowly with a broader peak width, and their sensitivities were reduced because of reduction in carrier gas flow rate at higher temperature. 12. In order to solve above problems, we tried to introduce a EPC system and a thinfilm-coated and high temperature (max. at 340°C) column DB-17ht (20 m × 0.32mm id, 0.15-µm film thickness). Of several GC conditions tested for this column, the three-ramp temperature programmes and EPC programs given in Subhead-

110

Kataoka et al.

Fig. 4. Standard chromatogram of the N(O,S)-isobutoxycarbonyl methyl ester derivatives of protein amino acids (containing 25 µg of asparagine, glutamine and cystine, and 5 µg of other amino acids) by FID-GC. GC conditions: column, DB-17 (15 m × 0.25 mm ID, 0.25-µm film thickness); column temperature, programmed at 30°C/min from 140°C to 290°C and then held for 5 min; injection and detector temperatures, 300°C; helium carrier gas and make-up gas flow rates, 1 mL/min and 30 mL/min, respectively; split ratio, 50:1. Peaks: 1 = alanine; 2 = glycine; 3 = valine; 4 = leucine; 5 = isoleucine; 6 = threonine; 7 = serine; 8 = proline; 9 = aspartic acid; 10 = glutamic acid; 11 = methionine; 12 = hydroxyproline; 13 = phenylalanine; 14 = asparagine; 15 = cysteine; 16 = glutamine; 17 = ornithine; 18 = lysine; 19 = histidine; 20 = tyrosine; 21 = tryptophan; 22 = cystine; IS = isonipecotic acid (from ref. 37, with permission). ing 3.2. were proved to give the most satisfactory separation of the protein and nonprotein amino acids. As shown in Fig. 5, the 21 protein amino acids and the 25 nonprotein amino acids could be well separated as single symmetrical peaks within 28 min on a DB-17ht capillary column, except for δ-HLys, which showed two peaks caused by the allo form present in the standard (41). 13. The reproducibility of sample injection, expressed as percent variation from the mean peak height ratio against the IS, was determined from three independent injections of the standard derivative mixture. For syringe manipulation, a hotneedle-injection technique (the needle inserted into injection zone was allowed to heat up for 3 s prior to the sample injection) was used to prevent the sample discrimination caused by incomplete vaporization and flashback into other parts of the inlet. A split-injection system was proved to be reproducible and the rela-


111

Fig. 5. Standard chromatogram obtained from the N(O,S)-isobutoxycarbonyl methyl ester derivatives of protein and nonprotein amino acids (containing 0.4 µg of asparagine and glutamine, and 0.2 µg of other amino acids) by NPD-GC. GC conditions: column, DB-17ht (20 m × 0.32 mm ID, 0.15-µm film thickness); column temperature, programmed at 4°C/min from 120 to 160°C, then programmed at 6°C/min from 160 to 200°C and at 10°C/min from 200 to 310°C; injection and detector temperatures, 320°C; inlet helium carrier gas flow rate, controlled 2 mL/min constant with EPC; make-up gas flow rate, 30 mL/min; split ratio, 10:1. Peaks: 1 = α-aminoisobutyric acid, 2 = alanine, 3 = glycine, 4 = α-aminobutyric acid, 5 = valine, 6 = β-alanine, 7 = β-aminobutyric acid, 8 = β-aminoisobutyric acid, 9 = norvaline, 10 = leucine, 11 = allo-isoleucine, 12 = isoleucine, 13 = norleucine, 14 = γ-aminobutyric acid, 15 = threonine, 16 = serine, 17 = proline, 18 = pipecolic acid, 19 = homoserine, 20 = δ-aminolevulinic acid, 21 = aspartic acid, 22 = thioproline, 23 = ε-aminocaproic acid, 24 = glutamic acid, 25 = methionine, 26 = hydroxyproline, 27 = α-aminoadipic acid, 28 = phenylalanine, 29 = α-aminopimelic acid, 30 = asparagine, 31 = p-aminobenzoic acid, 32 = 2,3-diaminopropionic acid, 33 = cysteine, 34 = glutamine, 35 = 2,4-diaminobutyric acid, 36 = homocysteine, 37 = methionine sulphone, 38 = ornithine, 39 = lysine, 40 = histidine, 41 = tyrosine, 42 = dhydroxylysine, 43 = tryptophan, 44 = cystathionine, 45 = cystine, 46 = homocystine, IS = 4-piperidinecarboxylic acid (from ref. 41, with permission). tive standard deviations ranged from 0.06 to 2.9%. On the other hand, a splitlessinjection system could introduced most of sample onto the column and increased sensitivity, but it was proved to reduce reproducibility because of poor refocusing (39). Particularly, the relative responses of high-boiling-point compounds, such as the derivatives of LTH, CTH, Cyt, and HCyt were remarkably reduced. Therefore, we adopted a hot-needle-injection technique in split mode as optimum GC injection technique.

112

Kataoka et al.

14. The calibration curves for protein amino acids by FID-GC were conducted using isonipecotic acid, which showed a similar behavior to other amino acids during the derivatization and was well separated from other amino acids on a chromatogram as the IS. A linear relationship was obtained with correlation coefficient being above 0.998 in the range 1–50 µg for Asn, Gln, and Cyt, 0.5–10 µg for Arg and 0.2–10 µg for other amino acids. The relative standard deviations for each amino acid in each point were 0.5–7.0% (n = 3). 15. The minimum detectable amounts of protein amino acids to give a signal three times as high as the noise under our FID-GC conditions were 0.2–4.0 ng as injection amounts. 16. The contents of Orn and 21 protein amino acids, except for Arg, were calculated from the directly derivatized samples without arginase treatment. On the other hand, Arg content was calculated by subtracting the amount of Orn obtained without arginase treatment from that obtained with arginase treatment (37). 17. NPD-GC is selective and sensitive for nitrogen-containing compounds and the application of this technique to the analysis of amino acids has been reported (8,12,48–52). The NPD-GC system described here was over 10–50 times more sensitive than the FID-GC. Particularly, the increase of sensitivity was remarkable for the nitrogen-rich amino acids such as DAPA, DABA, Orn, Lys, and His (40). 18. The calibration curves for protein and nonprotein amino acids by NPD-GC were conducted using 4-piperidinecarboxylic acid, which showed similar behavior to other amino acids during the derivatization and was well separated from other amino acids on a chromatogram as the IS. Various amounts of protein and nonprotein amino acids ranging from 0.02 to 2 µg were derivatized in the mixture, and aliquots representing 0.2–20 ng were injected into the NPD-GC system. As pointed previously (12), a significant curvature of the detector response was observed by NPD-GC, particularly at a low concentration of amino acid. Therefore, the calibration curves for each amino acid were constructed from both logarithmic plots of the peak height ratios and the amino acid amounts. A linear relationship was obtained with correlation coefficients being above 0.990, in the range 0.04–2 µg for Asn and Gln, and 0.02–1 µg for other amino acids (Table 1). 19. The minimum detectable amounts to give a signal-to-noise ratio of 3 under our NPD-GC conditions were 6–150 pg as injection amount (Table 1) (41). 20. In order to prevent the oxidation of Tyr and sulfur amino acids during acid hydrolysis, phenol was added to 6 M HCl at the concentration of 1%. In this hydrolysis, Trp is generally destroyed, Asn and Gln are quantitatively converted to Asp and Glu. 21. The FID-GC method developed was successfully applied to acid hydrolysate samples of proteins (10–100 µg) (37). Acid hydrolysis of proteins and peptides using a Pico-Tag workstation is not contaminated with reagent because of proceeding in the HCl vapor phase (43). 22. Typical chromatograms obtained from some proteins by FID-GC are shown in Fig. 6. The recoveries of total amino acids were 90–104%, and the reproducibility for each amino acid was satisfactory. The amino acid compositions of four well-characterized proteins, bovine serum albumin, hen ovalbumin, calf thymus


113

Fig. 6. Typical chromatograms obtained from acid hydrolysates of (A) bovine serum albumin (100 µg) and (B) collagen (100 µg) by FID-GC. GC conditions and peak number: see Fig. 4. Broken line shows the peak of arginine, which is analyzed as ornithine after arginase treatment (from ref. 37, with permission).

114

Kataoka et al.

Table 1 Linear Regression Data and Detection Limits for Protein and Nonprotein Amino Acids Regression linea Peak No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Amino acid* α-Aminoisobutyric acid Alanine Glycine α-Aminobutyric acid Valine β-Alanine β-Aminobutyric acid β-Aminoisobutyric acid Norvaline Leucine allo-Isoleucine Isoleucine Norleucine γ-Aminobutyric acid Threonine Serine Proline Pipecolic acid Homoserine δ-Aminolevulinic acid Aspartic acid Thioproline ε-Aminocaproic acid Glutamic acid

Abbreviation

Slope a

Intercept b

α-AIBA Ala Gly α-ABA Val β-Ala β-ABA β-AIBA NVaI Leu AIle Ile NLeu GABA Thr Ser Pro PCA HSer δ-ALA Asp TPro ε-ACA Glu

0.9981 1.0901 1.1637 1.0942 1.0465 1.2122 1.0833 1.1227 1.0645 1.0282 1.0163 1.0185 1.0670 1.1886 1.2551 1.4961 1.0113 1.0078 1.2982 1.13 13 0.9660 1.2683 1.0060 1.0266

0.9016 0.9456 1.0627 0.8731 0.7292 0.8740 0.8355 0.7655 0.7950 0.6375 0.6492 0.6627 0.6796 0.8782 0.7440 0.7727 0.8217 0.7320 0.6959 0.5799 0.6543 0.5937 0.5151 0.6256

Correlation Detection coefficient limit r (pg) 0.9967 0.9944 0.9952 0.9986 09988 0.9987 0.9990 0.9993 0.9991 0.9997 0.9989 0.9997 0.9999 0.9991 0.9995 0.9953 0.9996 0.9998 0.9992 0.9992 0.9997 0.9984 0.9994 0.9998

9 10 9 10 11 11 11 11 11 12 12 12 12 11 28 50 9 11 32 25 14 30 16 16 (continued)

histone, and calf skin collagen, determined by this method were good agreement with the literature values. By using this method, protein amino acid contents in serum could be rapidly and simply determined (38). 23. The NPD-GC method developed was successfully applied to the analysis of free amino acids in human urine (40,41) and serum (41) samples without prior cleanup procedures such as deproteinization, ion–exchange column chromatography, solid phase extraction, and subsequent eluate evaporation. 24. Typical chromatograms obtained from 25 µL of urine and 50 µL of serum by NPD-GC were shown in Fig. 7. In urine sample, hippuric acid was observed between TPro and Glu, but it was not overlapped with ε-ACA. Some unknown peaks were observed on the chromatogram, but 37 amino acids detected in the urine could be analyzed without any influence from coexisting substances.


115

Table 1 (continued) Linear Regression Data and Detection Limits for Protein and Nonprotein Amino Acids Regression linea Peak No. 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

Amino acid* Methionine Hydroxyproline α-Aminoadipic acid Phenylalanine α-Aminopimelic acid Asparagiine p-Aminobenzoic acid 2,3-Diaminopropionic acid Cysteine Glutamine 2,4-Diaminobutyric acid Homocysteine Methioninesulphone Ornithine Lysine Histidine Tyrosine δ-Hydroxylysine Tryptophan Cystathionine Cystine Homocystine

Correlation Detection coefficient limit r (pg)

Abbreviation

Slope a

Intercept b

Met Hyp α-AAA Phe α-APA Asn p-ABzA

1.3815 1. 1833 1.0597 0.9609 1.0683 1.3192 1.1047

0.2873 0.8384 0.5232 0.5718 0.4584 0.3684 0.6917

0.9971 0.9987 0.9992 0.9996 0.9983 0.9961 0.9980

135 16 24 15 28 56 20

DAPA Cys Gln DABA HCys Met-S Orn Lsy His Tyr δ-HLys Trp CTH Cyt HCyt

1.0236 1.1029 1.3119 1.0891 0.9898 1.0781 1.0267 1.0214 1.0178 1.4371 1.1698 1.2711 1.4594 1.2572 1.2690

1.0170 0.5802 0.1786 1.0254 0.1160 0.7250 1.1324 1.1195 1.0684 0.6400 0.6803 0.8663 0.7296 0.7120 0.4084

0.9984 0.9974 0.9968 0.9991 0.9931 0.9933 0.9982 0.9961 0.9963 0.9964 0.9934 0.9951 0.9936 0.9964 0.9901

9 24 150 10 60 25 6 6 7 64 34 24 30 40 70

alog y = a log x + b : y, peak height ratio against the I.S.; x, amount of each amino acid (µg); a, slope; b, intercept. Range: 0.04–2.0 µg for asparagine and glutamine; 0.02–1.00 µg for other amino acids (from ref. 41, with permission).

25. To confirm validity of NPD-GC method, known amounts of amino acids were spiked to human urine and serum, and their recoveries were calculated. As shown in Table 2, the overall recoveries of these amino acids were 83.5–112.0% and the relative standard deviations were 0.3–14.9% (n = 3). The quantitation limits of amino acids in urine and serum samples were approx 0.1–0.4 mg/mL. The intraassay C.V.s and inter-assay C.V.s for these samples throughout the overall procedure consisting of derivatization and GC analysis were 0.3–8.9% (n = 3) and 1.9–15.8% (n = 3), respectively.

116

Table 2 Recoveries of Amino Acids Added to Urine and Serum Samples Urine

Serum

Amount found (µg/mL)a

Added (µg/mL)

Nonaddition

α-AIBA Ala Gly α-ABA Val β-Ala β-ABA β-MBA NVal Lcu AIle Ile NLeu GABA Thr Ser Pro PCA HSer δ-ALA Asp TPro ε-ACA

2 20 20 2 2 2 2 20 2 2 2 2 2 2 20 20 2 2 2 2 2 2 2

8.61 ± 0.16 18.48 ± 1.41 36.49 ± 0.43 0.88 ± 0.07 3.69 ± 0.06 2.45 ± 0.09 NDb 108.2 ± 2.3 ND 3.80 ± 0.07 ND 1.33 ± 0.02 ND ND 11.13 ± 0.19 19.77 ± 0.99 0.79 ± 0.06 ND 0.88 ± 0.08 1.20 ± 0.07 0.80 ± 0.04 1.12 ± 0.07 ND

Amount found (µg/mL)

Addition

Recovery (%)

Added (µg/mL)

Nonaddition

Addition

10.59 ± 0.80 38.12 ± 2.17 57.59 ± 1.62 2.96 ± 0.07 5.59 ± 0.22 4.41 ± 0.20 2.13 ± 0.02 127.8 ± 3.0 2.15 ± 0.05 5.80 ± 0.12 2.03 ± 0.10 3.41 ± 0.08 2.16 ± 0.12 220 ± 007 29.80 ± 0.72 38.16 ± 1.96 2.80 ± 0.07 1.95 ± 0.06 2.91 ± 0.08 3.35 ± 0.23 2.75 ± 0.16 3.05 ± 0.26 1.88 ± 0.04

99.0 98.2 105.5 104.0 95.0 98.0 106.5 98.0 107.5 100.0 101.5 104.0 108.0 110.0 93.4 92.0 100.5 97.5 101.5 107.5 97.5 96.5 94.0

1 10 10 1 10 1 1 1 1 10 1 10 1 1 10 10 10 1 1 1 1 1 1

0.91 ± 0.08 25.68 ± 0.45 18.05 ± 0.35 3. 15 ± 0.13 29.33 ± 0.74 ND ND ND ND 17.52 ± 0.52 ND 10.11 ± 0.20 ND ND 19.85 ± 0.09 14.03 ± 0.19 27.65 ± 1.16 ND ND ND 1.10 ± 0.10 ND ND

1.89 ± 0.08 36.07 ± 2.09 28.23 ± 0.91 4.10 ± 0.04 39.75 ± 1.48 1.04 ± 0.12 1.10 ± 0.09 0.99 ± 0.03 1.01 ± 0.06 26.60 ± 1.68 1.12 ± 0.14 19.87 ± 1.01 0.98 ± 0.05 1.05 ± 0.05 29.34 ± 0.74 23.43 ± 0.76 37.65 ± 0.71 1.04 ± 0.04 0.97 ± 0.11 1 07 ± 0.03 209 ± 008 097 ± 0.12 1.07 ± 0.03

Recovery (%) 98.0 103.9 101.8 95.0 104.2 104.0 110.0 99.0 101.0 90.8 112.0 97.6 98.0 105.0 94.9 94.0 100.0 104.0 97.0 107.0 99.0 97.0 107.0

Kataoka et al.

Amino acid

2 2 2 2 2 2 4 2 2 2 40 2 20 2 2 20 20 20 20 20 2 20 2

2.28 ± 0.08 1.93 ± 0.22 ND 6.31 ± 0.02 5.65 ± 0.06 3.60 ± 0.12 16.40 ± 0.48 6.16 ± 0.08 ND 8.41 ± 0.13 59.54 ± 1.48 6.57 ± 0.05 16.68 ± 0.08 1.68 ± 0.14 6.60 ± 0.08 25.89 ± 0.54 158.0 ± 7.0 10.33 ± 0.67 10.71 ± 0.05 16.39 ± 0.22 5.01 ± 0.06 10.33 ± 0.40 6.44 ± 0.04

4.35 ± 0.14 3.60 ± 0.14 1.83 ± 0.16 8.25 ± 0.30 7.75 ± 0.12 5.51 ± 0.30 20.40 ± 0.22 8.28 ± 0.16 2.08 ± 0.18 10.39 ± 0.46 97.02 ± 6.58 8.43 ± 0.53 34.87 ± 2.36 3.55 ± 0.26 8.49 ± 0.13 45.17 ± 1.61 178.2 ± 9.5 31.76 ± 0.76 28.73 ± 0.94 36.01 ± 2.39 7.04 ± 0.31 30.05 ± 1.24 8.27 ± 0.05

103.5 83.5 91.5 97.0 105.0 95.5 100.0 106.0 104.0 99.0 93.7 93.0 91.0 93.5 94.5 96.4 101.0 107.2 90.1 98.1 101.5 98.6 91.5

10 I 1 1 10 1 2 1 1 1 20 1 1 1 10 10 1 1 1 1 1 1 1

13.86 ± 0.52 1.00 ± 0.03 1.95 ± 0.07 ND 10.94 ± 0.39 ND 7.52 ± 0.62 ND ND 0.94 ± 0.04 100.2 ± 2.8 0.10 ± 0.01 ND 0.51 ± 0.06 8.97 ± 0.37 34.15 ± 1.37 3.55 ± 0.29 8.49 ± 0.23 0.54 ± 0.05 6.47 ± 0.30 3.32 ± 0.10 0.51 ± 0.04 ND

23.51 ± 0.54 1.86 ± 0.04 2.95 ± 0.21 1.04 ± 0.02 21.06 ± 0.65 0.99 ± 0.02 9.66 ± 0.38 1.09 ± 0.04 0.99 ± 0.08 1.87 ± 0.24 119.1 ± 3.2 1.07 ± 0.04 0.97 ± 0.08 1.48 ± 0.22 18.55 ± 1.23 44.64 ± 2.85 4.49 ± 0.52 9.48 ± 0.63 1.46 ± 0.05 7.38 ± 0.19 4.23 ± 0.36 1.55 ± 0.10 1.08 ± 0.02

96.5 86.0 100.0 104.0 101.2 99.0 107.0 109.0 99.0 93.0 94.5 97.0 97.0 97.0 95.8 104.9 94.0 99.0 92.0 91.0 91.0 104.0 108.0


Glu Met Hyp α-AAA Phe α-APA Asn p-ABzA DAPA Cys Gln DABA HCys Met-S Orn Lsy His Tyr δ-HLys Trp CTH Cyt HCyt

± SD (n = 3). detectable (from ref. 41, with permission).

aMean bNot

117

118

Kataoka et al.

Fig. 7. Typical chromatograms obtained from (A) urine (25 µL) and (B) serum (50 µL) samples by NPD-GC. GC conditions and peak number: see Fig. 5.


119

References 1. Cruickshank, P. A. and Sheehan, J. C. (1964) Gas chromatographic analysis of amino acids as N-trifluoroacetyl amino acid methyl esters. Anal. Chem. 36, 1191–1197. 2. Darbre, A. and Islam, A. (1968) Gas-liquid chromatography of trifluoroacetylated amino acid methyl esters. Biochem. J. 106, 923–925. 3. Gamerith, G. (1983) Gas-liquid chromatographic determination of N(O, S)trifluoroacetyl n-propyl esters of protein and non-protein amino acids. J. Chromatog. 256, 267–281. 4. Gehrke, C. W., Kuo, K. C., Kaiser, F. E., and Zumwalt, R. W. (1987) Analysis of amino acids by gas chromatography as the N-trifluoroacetyl n-butyl esters. J. Assoc. Off. Anal. Chem. 70, 160–170. 5. Zumwalt, R. W., Desgres, J., Kuo, K. C., Pautz, J. E., and Gehrke, C. W. (1987) Amino acid analysis by capillary gas chromatography. J. Assoc. Off. Anal. Chem. 70, 253–262. 6. Singh, A. K. and Ashraf, M. (1988) Analysis of amino acids in brain and plasma samples by sensitive gas chromatography-mass spectrometry. J. Chromatog. 425, 245–255. 7. MacKenzie, S. L. and Tenaschuk, D. (1974) Gas-liquid chromatography of Nheptafluorobutyryl isobutyl esters of amino acids. J. Chromatog. 97, 19–24. 8. Chauhan, J., Darbre, A., and Catlyle, R. F. (1982) Determination of urinary amino acids by means of glass capillary gas-liquid chromatography with alkali-flame ionisation detection and flame ionisation detection J. Chromatog. 227, 305–321. 9. Schneider, K., Neupert, M., Spiteller, G., Henning, H. V., Matthaei, D., and Scheler, F. (1985) Gas chromatography of amino acids in urine and haemofiltrate. J. Chromatog. 345, 19–31. 10. Labadarios, D., Shephard, G. S., Botha, E., Jackson, L., Moodie, I. M., and Burger, J. A. (1986) Determination of plasma amino acids by gas chromatography. J. Chromatog. 383, 281–295. 11. MacKenzie, S. L. (1987) Gas chromatographic analysis of amino acids as the Nheptafluorobutyryl isobutyl esters. J. Assoc. Off. Anal. Chem. 70, 151–160. 12. Philpott, M. F. and Van, M. J., der Merwe (1991) Non-linear calibration of a nitrogen-phosphorus detector for the GC determination of amino acids. Chromatographia 31, 500–504. 13. Yeung, J. M., Baker, G. B., and Coutts, R. T. (1986) Simple automated gas chromatographic analysis of amino acids and its application to brain tissue and urine. J. Chromatog. 378, 293–304. 14. Gehrke, C. W. and Leimer, K. (1971) Trimethylsilylation of amino acids. Derivatization and chromatography. J. Chromatog. 57, 219–238. 15. Biermann, C. J., Kinoshita, C. M., Marlett, J. A., and Steele, R. D. (1986) Analysis of amino acids as tert.-butyldimethylsilyl derivatives by gas chromatography. J. Chromatog. 357, 330–334. 16. Mawhinney, T. P., Robinett, R. S. R., Atalay, A., and Madson, M. A. (1986) Analysis of amino acids as their tert.-butyldimethylsilyl derivatives by gas chromatography and mass spectrometry. J. Chromatog. 358, 231–242.

120

Kataoka et al.

17. MacKenzie, S. L., Tenaschuk, D., and Fortier, G. (1987) Analysis of amino acids by gas-liquid chromatography as tert.-butyldimethylsilyl derivatives. Preparation of derivatives in a single reaction. J. Chromatog. 387, 241- 253. 18. Early, R. J., Thompson, J. R., Sedgwick, G. W., Kelly, J. M., and Christopheason, R. J. (1987) Capillary gas chromatographic analysis of amino acids in blood and protein hydrolysates as tert.-butyldimethylsilyl derivatives. J. Chromatog. 416, 15–23. 19. Goh, C. J., Craven, K. G., Lepock, J. R., and Dumbroff, E. B. (1987) Analysis of all protein amino acids as their tert.-butyldimethylsilyl derivatives by gas-liquid chromatography. Anal. Biochem. 163, 175–181. 20. Woo, K. L. and Lee, D. S. (1995) Capillary gas chromatographic determination of proteins and biological amino acids as N(O)-tert.-butyldimethylsilyl derivatives. J. Chromatog. 665, 15–25. 21. Makita, M., Yamamoto, S., and Kono, M. (1976) Gas-liquid chromatographic analysis of protein amino acids as N-isobutoxycarbonylamino acid methyl esters. J. Chromatog. 120, 129–140. 22. Makita, M., Yamamoto, S., Sakai, K., and Shiraishi, M. (1976) Gas-liquid chromatography of the N-isobutoxycarbonyl methyl esters of non-protein amino acids. J. Chromatog. 124, 92–96. 23. Yamamoto, S., Kiyama, S., Watanabe, Y., and Makita, M. (1982) Practical gasliquid chromatographic method for the determination of amino acids in human serum. J. Chromatog. 233, 39–50. 24. Makita, M., Yamamoto, S., and Kiyama, S. (1982) Improved gas-liquid chromatographic method for the determination of protein amino acids. J. Chromatog. 237, 279–284. 25. Oh, C.-H., Kim, J.-H., Kim, K.-R., Brownson, D. M., and Mabry, T. J. (1994) Simultaneous gas chromatographic analysis of non-protein and protein amino acids as N(O, S)-isobutoxycarbonyl tert.-butyldimethylsilyl derivatives. J. Chromatog. 669, 125–137. 26. Husek, P. (1991) Rapid derivatization and gas chromatographic determination of amino acids. J. Chromatog. 552, 289–299. 27. Cao, P. and Moini, M. (1997) Quantitative analysis of fluorinated ethylchloroformate derivatives of protein amino acids and hydrolysis products of small peptides using chemical ionization gas chromatography-mass spectrometry. J. Chromatog. A, 759, 111–117. 28. Husek, P., Rijks, J. A., Leclercq, P. A., and Cramers, C. A. (1990) Fast esterification of fatty acids with alkyl chloroformates. Optimization and application in gas chromatography. J. High Resolut. Chromatogr. 13, 633–638. 29. Husek, P. (1993) Capillary GC analysis of biogenic amines, their precursors and catabolytes after fast derivatization with ethyl chloroformate. J. Microcol. Sep. 5, 101–103. 30. Husek, P. (1995) Simultaneous profile analysis of plasma amino and organic acids by capillary gas chromatography. J. Chromatog. B, 669, 352–357. 31. Rattenbury, J. M. (ed.) (1981) Amino Acid Analysis. Ellis Horwood, Chichester, U.K.


121

32. Zumwalt, R. W., Kuo, K. C. T., and Gehrke, C. W. (eds.) (1987) Amino Acid Analysis by Gas Chromatography, vol. I–III. CRC, Boca Raton, FL. 33. Clement, R. E. (ed.) (1990) Gas Chromatography. Biochemical, Biomedical and Clinical Applications. Wiley, New York. 34. MacKenzie, S. L. (1981) Recent developments in amino acid analysis by gas-liquid chromatography. Methods Biochem. Anal. 27, 1–88. 35. Labadarios, D., Moodie, I. M., and Shephard, G. S. (1984) Gas chromatographic analysis of amino acids in physiological fluids: a critique. J. Chromatog. 310, 223–231. 36. Walker, V. and Mills, G. A. (1995) Quantitative methods for amino acid analysis in biological fluids. Ann. Clin. Biochem. 32, 28–57. 37. Matsumura, S., Kataoka, H., and Makita, M. (1995) Capillary gas chromatographic analysis of protein amino acids as their N(O, S)-isobutoxycarbonyl methyl ester derivatives. Biomed. Chromatogr. 9, 205–210. 38. Matsumura, S., Kataoka, H., and Makita, M. (1996) Determination of amino acids in human serum by capillary gas chromatography. J. Chromatog. B 681, 375–380. 39. Kataoka, H., Matsumura, S., Koizumi, H., and Makita, M. (1997) Rapid and simultaneous analysis of protein and non-protein amino acids as N(O, S)isobutoxycarbonyl methyl ester derivatives by capillary gas chromatography. J. Chromatog. A, 758, 167–173. 40. Kataoka, H. (1997) Selective and sensitive determination of protein and non-protein amino acids by capillary gas chromatography with nitrogen-phosphorus selective detection. Biomed. Chromatogr. 11, 154–159. 41. Kataoka, H., Matsumura, S., and Makita, M. (1997) Determination of amino acids in biological fluids by capillary gas chromatography with nitrogen-phosphorus selective detection. J. Pharm. Biomed. Anal. 15, 1271–1279. 42. Schlenk, H. and Gellerman, J. L. (1960) Esterification of fatty acids with diazomethane on a small scale. Anal. Chem. 32, 1412–1414. 43. Bidlingmeyer, B. A., Cohen, S. A., and Tarvin, T. L. (1984) Rapid analysis of amino acids using pre-column derivatization. J. Chromatog. 336, 93–104. 44. Kataoka, H., Tanaka, H., Fujimoto, A., Noguchi, I., and Makita, M. (1994) Determination of sulphur amino acids by gas chromatography with flame photometric detection. Biomed. Chromatogr. 8, 119–124. 45. Makita, M., Yamamoto, S., Katoh, A., and Takashita, Y. (1978) Gas chromatography of some simple phenols as their O-isobutoxycarbonyl derivatives. J. Chromatog. 147, 456–458. 46. Makita, M., Yamamoto, S., Miyake, M., and Masamoto, K. (1978) Practical gas chromatographic method for the determination of urinary polyamines. J. Chromatog. 156, 340–345. 47. Yamamoto, S., Kakuno, K., Okahara, S., Kataoka, H., and Makita, M. (1980) Gas chromatography of phenolic amines, 3-methoxycatecholamines, indoleamines and related amines as their N, O-ethoxycarbonyl derivatives. J. Chromatog. 194, 399–403. 48. Butler, M. and Darbre, A. (1974) Determination of amino acids by gas-liquid chromatography with the nitrogen-sensitive thermionic detector. J. Chromatog. 101, 51–56.

122

Kataoka et al.

49. Adams, R. F., Vandemark, F. L., and Schmidt, G. J. (1977) Ultramicro GC determination of amino acids using glass open tubular columns and a nitrogen-selective detector. J. Chromatog. Sci. 15, 63–68. 50. Frank, H., Vujtovic-Ockenga, N., and Rettenmeier, A. (1983) Amino acid determination by capillary gas chromatography on chirasil-val. Enantiomer labelling and nitrogen-selective detection. J. Chromatog. 279, 507–514. 51. MacKenzie, S. L. (1986) Amino acid analysis by gas-liquid chromatography using a nitrogen-selective detector. J. Chromatog. 358, 219–230. 52. Buser, W. and Erbersdobler, H. F. (1988) Gas chromatographic determination of amino acids with nitrogen-selective detection. Zeitsh. Lebens.-Unter. Forsch. 186, 509–513.

Analysis of Blood Amino Acids by HPLC Method

123

10 Measurement of Blood Plasma Amino Acids in Ultrafiltrates by High-Performance Liquid Chromatography with Automatic Precolumn O-Phthaldialdehyde Derivatization Hua Liu 1. Introduction Amino acid analysis is an important technique that has many applications in biochemical, pharmaceutical, and biomedical fields. Profiling of plasma amino acids is also of great interest in clinical practice. Many diseases are known to be associated with disorders in amino acid metabolism. The analysis of amino acids offers the possibility of genetic prevention in both premarital and prenatal stages for those diseases resulting from inborn errors of metabolism. The plasma amino acid pattern has been also used to follow the course of prolonged dietary treatment (1,2). For nearly 40 years, amino acid separations have been carried out mainly by an amino acid analyzer by means of ion–exchange chromatography and detected after postcolumn derivatization (1,3). The use of this analyzer has been widely advocated in the past, especially in routine application because of its high reliability (4). However, these analyzes are somewhat laborious, costly, time-consuming, and usually performed on dedicated instruments (1,5,6). In recent years, methods employing precolumn derivatization of amino acids, combined with reversed-phase (RP) high-performance liquid chromatography (HPLC) have gained increasing importance and have partially replaced the classical amino acid analyzer (6). Publications have demonstrated the usefulness of this technique for the determination of amino acids in physiological fluids, and the results achieved by HPLC methods compared favorably with those

From: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

123

124

Liu

obtained with the amino acid analyzer (6–14). In comparison with the postcolumn derivatization and ion–exchange chromatographic amino acid analyzers, precolumn derivatization and HPLC methods has the advantages of reducing analysis times, enhancing sensitivity and flexibility, and lowering cost of instrumentation and maintenance (7–16). Among the methods of precolumn derivatization, O-phthaldialdehyde (OPA) has become the most popular derivatization reagent. In the presence of a thiol compound and at alkaline pH, OPA reacts with primary amino acids and forms highly fluorescent isoindole derivatives. This derivatization procedure is relatively easy and the reaction occurs rapidly at room temperature and in aqueous solution. No laborious purification procedures are required. The OPA derivatives are less polar than the original amino acids and can be well separated from each other by RP-HPLC (17,18). Moreover, the reagent itself does not fluoresce and consequently produces no interfering peaks. One disadvantage of OPA derivatization has been the lack of stability of the OPA adducts when 2-mercaptoethanol (2-ME) is used as a sulfhydryl reagent (17,19). The reaction products are not stable and have a short half-life, possibly because of a spontaneous intramolecular rearrangement, with sulfur being displaced by oxygen from the ethanolic portion (19). Consequently, time differences between the reaction and injection during a manual procedure may cause significant errors in quantitation. An automated on-line OPA derivatization procedure is described in this chapter. By using an autoinjector, the exact time of each step from the beginning of reaction to injection can be controlled according to an injector program. The precision of the reaction time and sample volume in an automatic procedure eliminates the human errors that may occur during a manual derivatization. In addition, 2-ME was replaced by 3-mercaptopropionic acid (3-MPA) in our experiment with considerable improvement in stability because the stability of isoindoles formed. The other advantages of 3-MPA are that it is nonvolatile and less toxic because of its carboxylic moiety (20). Another disadvantage of OPA derivatization is that it reacts only with primary amines, so secondary amino acids (imino acids) are not detected (19,21). 9-fluorenylmethyl-chloroformate (FMOC-Cl) is a highly reactive reagent that has been used as an amino-protective group in peptide synthesis. In 1983, Einarsson applied FMOC-Cl to a precolumn derivatization of amino acids (22). It was as sensitive as OPA, but reacted with secondary, as well as primary amine. The derivatives produced were stable and highly fluorescent. This method has been used to measure hydroxyproline, sarcosine and proline in serum, cerebrospinal fluid, and urine (22,23). FMOC-Cl itself is fluorescent and may obscure some amino acid peaks. Therefore, excess FMOC-Cl has to be removed by pentane extraction or reacted with hydrophobic amine 1-amin-


125

oadamantane (ADAM) to form an amine-FMOC complex (24). We obviated this problem by using a double-derivative technique (FMOC-Cl is incorporated to this automatic derivatization procedure as a second reagent for the derivatization of secondary amino acids) and using both photodiode array and programmable fluorescence detectors. This procedure has the advantage of different absorption and fluorescent spectra of OPA and FMOC-Cl derivatives. Therefore, both primary and secondary amino acids can be detected simultaneously. Although there are many techniques available for the analysis of amino acids, and the sample can be analyzed quickly, accurately and sensitively, several precautions need to be taken in order to obtain reliable data on the concentration of plasma amino acids. These include the sample collection, centrifugation, storage conditions, and the deproteinization method. Deproteinization is one of the major problems in the analysis of amino acids in physiological fluid (25). The whole plasma contains soluble peptides and proteins that should be removed from the sample. Otherwise, these substances will clog the chromatographic column, increase instrumental backpressure, and interfere with separation. The method used to prepare the plasma and to remove the plasma protein has a marked effect on the final results (1,3). The most widely used method of deproteinization is precipitation with 5-sulfosalicylic acid (SSA) followed by centrifugation to remove the precipitated protein (1). We have not had success in using SSA as the deproteinization agent for the analysis of plasma amino acids by an HPLC method with automatic precolumn OPA/3-MPA and FMOCCl derivatization (11). When this method was used for the analysis of plasma samples deproteinized by SSA, several major problems were encountered. First, the yield of the derivatization was low in the SSA supernatant. This is probably because the strong acidic nature of SSA inhibits the formation of OPA-amino acid derivatives, which require an alkaline pH. Second, the large peak of deproteinizing agent SSA superimposed the first three amino acid peaks of the chromatogram (O-phospho-l-serine (OPS), Asp and Glu). The SSA peaks were higher than 4000, 2000, and 300 mAu at the UV sample wavelengths 230, 260, and 338 nm, respectively. The third problem was the adverse affects of SSA supernatant on the separation and quantitation of other amino acids. Other investigators (11,18) have observed similar problems. When ethanol or methanol was used for the deproteinization, the sample was diluted and some of plasma amino acids with lower levels (OPS, Asp, Glu, AABA, and Trp) became undetectable by the UV detector at 338 nm. Other problems were the high level of organic solvent in the injected solution resulting in broad peaks in the early part of the chromatogram, and the increased volatility of the sample made it difficult to store for long periods of time (11).

126

Liu

Ultrafiltration has the advantage of achieving a protein-free sample without adding chemical agents, thus keeping the sample close to a physiological state. It is an attractive alternative to equilibrium dialysis because of the ease and speed with which it can be accomplished. Another potential benefit of ultrafiltration is that the platelets and leukocytes are removed from the plasma, and thus the contamination of amino acids from these blood components can be eliminated. The ultrafiltration methods are not widely applied in the ion–exchange chromatography of amino acids because, for unknown reasons, they decrease the retention time during chromatographic separation and consequently leads to distorted separations of critical pairs of amino acids (1,3). Ultrafiltration has been chosen for the preparation of protein-free sample in the analysis of amino acid by HPLC methods with precolumn derivatization (18,26,27). The results indicate that ultrafiltration of plasma may replace chemical deproteinization in the HPLC analysis of free amino acids. We investigated several factors that may have an influence on the final results of the ultrafiltration (11). These conditions were then standardized in our procedure. Satisfactory results were achieved for the analysis of plasma amino acids by the automatic precolumn OPA/3-MPA and FMOC-Cl derivatization and RPHPLC method. It appears that the combination of: (1) automatic on-line derivatization (2) the improvement in the stability of the reaction by using 3-MPA (3) the use of OPA and FMOC-Cl as dual derivatizing reagents (4) the optimized gradient elution program (5) and the simple mobile phase composition render this method suitable for the quantitative analysis of plasma amino acids in a clinical laboratory. This procedure can yield more rapid and sensitive results than that of classic amino acid analyzer, with comparable accuracy and precision. We have used this method for the determination of plasma amino acids in more than 3000 plasma samples during a period of 10 yr with satisfactory results (see Notes 1 and 2). 2. Materials

2.1. Equipment 1. A Hewlett-Packard HP 1090 M series HPLC system (see Note 3). The system consists of a DR 5 solvent delivery system with 3 solvent channels, a variable volume autoinjector and an autosampler, an HP 1040A photo-diode array UV detector, and HP 1046 programmable fluorescence detector (see Note 4). 2. An HP 79994A analytical workstation for data processing. 3. Two HP Hypersil-ODS 5-µm columns (100 × 2.1 mm). 4. One guard column (20 × 2.1 mm) (see Note 5).


127

2.2. Reagents Water, methanol, and acetonitrile were HPLC grade (Curtin Matheson Scientific, Inc., Houston, TX). Other chemicals used were analytical grade including sodium acetate, glacial acetic acid, boric acid, sodium hydroxide, O-phthaldialdehyde, 3-mercaptopropionic acid, and 9-fluorenylmethyl-chloroformate (all from Sigma Chemical Company, St. Louis, MO). 2.3. Solutions 1. Standard solution: An amino acid standard solution containing 29 amino acids was prepared by adding crystalline O-phospho-L-serine (OPS), reduced glutathione (GSH), asparagine, glutamine, citrulline, taurine, α-amino-N-butyric acid (AABA), tryptophan, ornithine, hydroxy-proline (Hyp), and sarcosine to a 50% methanol solution. Then this solution (containing 11 amino acids) was mixed with a commercial AA-S-18 Amino Acid Standard Solution (from Sigma). NorValine (N-Val) can be used as an internal standard (United States Biochem Corp, Cleveland, OH). The concentration of working standard was 250 µmol/L for each amino acid. The mixture solution was kept in –76°C (see Note 6). 2. OPA derivatization reagent: The OPA derivatization reagent was prepared by dissolving 3 mg of OPA in 50 µL of methanol, adding 450 mL of sodium borate buffer (0.5 mol/L, pH 10.2) and 5 µL of 3-MPA. Borate buffer was prepared from 0.5 M boric acid solution adjusted to pH 10.2 with 5 M sodium hydroxide solution. This OPA solution was placed in an amber crimp top vial with a silicone rubber PTFE-coated cap and kept in the dark at –20°C. Fresh solution was prepared each week. 3. FMOC derivatization reagent: FMOC-Cl solution was prepared by dissolving 1.29 mg of FMOC-Cl in 1 mL of acetonitrile and stored at –20°C (see Note 7). 4. Mobile phase solution: The sodium acetate buffer (0.1 M) in mobile phase was prepared by dissolving sodium acetate in HPLC-grade water and titrating to pH 6.8 with glacial acetic acid. The buffer was then diluted to 0.015 M and 0.01 M for mobile phase solutions A and C, respectively (see Note 8). The mobile phases were filtered by passing through a 0.45-µm Durapore membrane filter (Millipore Inc., Milford, MA) and continuously degassed by helium. 5. Blood sample: Blood samples were collected from subjects into a heparinized vacutainer tube (see Note 9). The plasma samples were ultrafiltrated by using the Centrifree System (Amicon, Beverly, MA).

3. Methods 1. Collect 1.5 mL of venous blood samples from subjects by venipuncture between 7 AM to 9 AM (see Note 10) after overnight fasting and put into a vacutainer tube containing heparin (see Note 11). Centrifuge the blood samples at 2000g for 15 min at 10°C (see Note 12) and remove the plasma. 2. Ultrafiltrate the heparinized plasma samples by using the commercially available Centrifree System (see Notes 13 and 14). During the ultrafiltration, the sample is

128

Liu

Table 1 Injector Programa Line# 1 2 3 4 5 6 7 8 9 10 aReprinted

Function Draw Draw Draw Draw Mix Draw Draw Mix Wait Inject

: : : : : : : : :

Amount 0.0 µL from 2.5 µL from 0.0 µL from 2.5 µL from 5.0 µL cycles 0.0 µL from 1.0 µL from 6.0 µL cycles 2.5 min

Vial No. : : : : 2 : : 2

Vial#:4 Vial#:5 Vial#:4 Vial#:X Vial#:4 Vial#:8

Reagent (Water) (OPA) (Sample)

(FMOC-Cl)

from ref. 8, p. 3326 by courtesy of Marcel Dekker, Inc.

Table 2 Time Table for Gradient Elution Time (Minute) 0.05 15.00 18.50 22.00 25.00 30.00

Solvent A%

B%

C%

100 60 57.5 45 0 0

0 40 42.5 55 0 0

0 0 0 0 100 100

Solvent A: 0.015 M NaAc buffer (pH 6.8). Solvent B: Methanol. Solvent C: 0.010 M NaAc buffer (pH 6.8). Reprinted from ref. 8, p. 3327 by courtesy of Marcel Dekker, Inc.

deproteinized by filtration of plasma through the ultrafiltration membrane. The protein is retained by the membrane, whereas the ultrafiltrate (containing free amino acids) pass through and collect in the filtrater cup. Put a volume of 0.4 mL of plasma in the sample reservoir, and then place the device in a centrifuge with a 45° fixed-angle rotor (see Note 15). About 80 µL of ultrafiltrates should be collected after centrifugation at 1000g for 15 min. Store the ultrafiltrates at –80°C until analyzed (see Note 16). 3. To perform the automatic precolumn derivatization procedure using an injector program (Table 1), place the OPA-reagent at vial number 5 and the FMOC-Cl reagent at vial number 8. First, 2.5 µL of the OPA reagent is drawn into the sample loop. Then, 2.5 µL of sample is drawn into the loop and mixed with the OPA reagent. Finally, 1 µL of FMOC-CL reagent is drawn and mixed in the sample


4.

5.

6.

7.

129

loop. After waiting 2.5 min for the reaction, a total 6 µL of sample and reagents mixture is injected into the column for gradient elution (see Note 17). After drawing from sample or reagents, the needle of the injector should always be dipped into 1 mL of water (vial number 4) for cleaning. The mobile phases are 0.015 M sodium acetic buffer (pH 6.8, for solvent A), methanol (solvent B) and 0.015 M of sodium acetic buffer (pH 6.8, for solvent C) (see Notes 18 and 19). The separation of amino acids in both the standard solution containing 29 amino acids or plasma samples is carried out by a gradient elution according to a chromatographic time-table (Table 2). The flow-rate is 0.3 mL/min and the stop time is 30 min after the injection. Place two HP HypersilODS 5-µm columns (100 × 2.1-mm id) in series in a thermostatically controlled column compartment preceded by a guard column (20 × 2.1-mm id). The column temperature should be maintained at 40°C for the separation of amino acids (see Note 20). For the detection of amino acid derivatives, set the photo-diode array detector at three sample wavelengths: 338, 266, and 230 nm with bandwidths of 10, 4, and 4 nm, respectively (see Note 21). The reference wavelength is 550 nm with a bandwidth of 100 nm. The initial parameters for the fluorescence detector are excitation wavelength (Ex) 230 nm and emission wavelength (Em) 450 nm. Twenty minutes after the injection, change the Ex and Em to 260 nm and 315 nm, respectively, for the determination of secondary amino acids proline and hydroxproline (see Note 22). Amino acid peaks are identified with reference to retention times of standard amino acids injected. Coinjection of standard amino acids and the plasma samples may be needed to identify the amino acid peaks in some samples. Chromatograms of an amino acids standard mixture and a representative plasma sample (see Note 11) are demonstrated in Figs. 1 and 2. These chromatograms showed a satisfactory separation of 29 primary and secondary amino acids. The precision of analysis is observed from the reproducibility of the peak areas of eight consecutive injections of 29 amino acid mixture solution. The results are listed in Table 3. The coefficients of variation for peak areas ranged from 0.78% to 2.92%, with a mean of 1.73% ± 0.67% SD (see Note 23). The high precision of this method would allow analysis without an internal standard for quantitation. The amino acid standard should be first analyzed twice for calibration and thereafter every eighth analysis in an automated series.

4. Notes 1. The levels of 25 plasma amino acids from 75 boys and 85 girls are listed in Table 4. In general, amino acid levels of boys were higher than that of girls. The mean values of aspartic acid, methionine, isoleucine, and hydroxy-proline in boys were significantly higher than that in girls (p < 0.01). The girls had significant higher level of histidine than boys (p < 0.001). Compared with boys below 6 yr old (n = 23), boys over 6 yr old (n = 52) showed significant higher value of glycine, threo-

130

Liu

Fig. 1. Typical chromatograms showing the separation of 29 OPA/3-MPA and FMOC-Cl derivatized standard amino acids (500 µmol/L). (A) Signal from UV detection at 338 nm. (B) Signal from Fluorescence detection. For chromatographic conditions, see Subheadings 2 and 3. Peaks: 1 = O-phospho-L-serine, 2 = Aspartic Acid, 3 = Glutamic Acid, 4 = Glutathione (reduced), 5 = Asparagine, 6 = Serine, 7 = Glutamine, 8 = Glycine, 9 = Threonine, 10 = Histidine, 11 = Cystine, 12 = Citrulline, 13 = Taurine, 14 = Alanine, 15 = Arginine, 16 = Tyrosine, 17 = Alpha-amino-N-butyric Acid, 18 = Methionine, 19 = Valine, 20 = Nor-Valine, 21 = Tryptophan, 22 = Phenylalanine, 23 = Isoleucine, 24 = Ornithine, 25 = Leucine, 26 = Lysine, 27 = Hydroxy-proline, 28 = Sarcosine, 29 = Proline. Reprinted from ref. 8, p. 3328 by courtesy of Marcel Dekker, Inc.


131

Fig. 2. Chromatograms of a representative plasma sample under the identical experimental conditions as in Fig. 1. (A) Signal from UV detection at 338 nm. (B) Signal from fluorescence detection. Peak numbers are identified in Fig. 1A and B. Reprinted from ref. 8, p. 3329 by courtesy of Marcel Dekker, Inc. nine, histidine, taurine, alanine, arginine, tyrosine, valine, phenylalanine, isoleucine, ornithine and leucine. For girls over 6 yr old (n = 56), only threonine, histidine, and alanine were significantly higher than that for girls below 6 yr old (n = 29) (8).

132

Liu

Table 3 Reproducibility of Peak Areas for Standard Amino Acids Amino Acida OPS Asp Glu GSH Asn Ser Gln Gly Thr His Cys Cit Tau Ala Arg Tyr AABA Met Val N-Val Trp Phe Ile Orn Leu Lys Hyp Sar Pro

Mean (n = 8) 527.6 528.1 545.4 532.4 581.9 694.3 676.1 498.8 536.4 106.0 170.8 637.7 582.4 589.0 554.4 553.7 707.5 548.4 660.3 829.4 638.4 535.1 646.1 540.6 545.1 587.8 1469 2427 645.9

SD

CV%

5.47 4.13 15.49 13.69 16.58 8.96 8.99 9.97 10.78 1.59 3.94 4.88 5.76 5.22 7.37 8.14 13.78 10.52 7.69 12.16 16.49 11.55 18.51 15.80 10.01 9.54 28.91 20.85 8.86

1.03 0.78 2.84 2.57 2.84 1.29 1.32 1.99 2.00 1.50 2.30 0.76 0.98 0.88 1.32 1.47 1.94 1.91 1.16 1.46 2.58 2.15 2.86 2.92 1.83 1.62 1.91 0.85 1.37

aOPS

= O-phospho-L-serine, Asp = Aspartic Acid, Glu = Glutamic Acid, GSH = Glutathione (reduced), Asn = Asparagine, Ser = Serine, Gln = Glutamine, Gly = Glycine, Thr = Threonine, His = Histidine, Cys = Cystine, Cit = Citrulline, Tau = Taurine, Ala = Alanine, Arg = Arginine, Tyr = Tyrosine, AABA = Alpha-amino-N-butyric Acid, Met = Methionine, Val = Valine, N-Val = Nor-Valine, Trp = Tryptophan, Phe = Phenylalanine, Ile = Isoleucine, Orn = Ornithine, Leu = Leucine, Lys = Lysine, Hyp = Hydroxy-proline, Sar = Sarcosine, Pro = Proline.

2. The plasma amino acid levels of 160 children obtained by this method showed a good agreement with the reference values reported previously by using ion–exchange chromatography (28). The findings about the difference between amino acid levels of boys and girls and between the children with different age were


133

Table 4 Free Plasma Amino Acid Levels (mmol/L) of 160 Children Girls (n = 85) Amino Acida Asp Glu Asn Ser Gln Gly Thr His Cys Cit Tau Ala Arg Tyr AABA Met Val Trp Phe Ile Orn Leu Lys Hyp Pro

Mean 6.19 39.28 53.99 126.30 532.80 242.50 142.40 96.09c 88.96 24.29 41.57 423.30 95.05 74.88 18.23 23.88 221.80 5.67 64.94 66.36 42.11 137.90 130.00 18.96 153.00

±

SD 1.86 13.95 10.16 23.57 82.00 47.29 27.86 16.51 14.91 4.10 7.73 82.15 18.18 15.45 3.52 2.60 39.27 2.73 11.15 15.35 12.76 25.92 25.99 4.20 42.24

Boys (n = 75) Mean 7.62b 48.23 56.53 133.50 540.70 262.60 146.20 85.07 87.21 26.00 44.35 425.10 100.60 73.99 18.66 25.17b 235.20 6.38 65.32 74.28b 45.03 144.30 126.10 21.34b 149.90

±

SD 2.13 17.84 10.92 28.79 93.75 59.35 28.87 16.95 15.29 4.15 11.53 96.85 19.52 13.51 3.93 3.65 49.11 3.33 5.96 19.63 14.53 32.03 26.91 5.65 45.66

aAmino

acid abbreviations as in Table 3. higher than girls (p < 0.01). cSignificantly higher than boys (p < 0.001). Reprinted from ref. 8, p. 3332 by courtesy of Marcel Dekker, Inc. bSignificantly

also similar to the results observed by Armstrong and Stave (28,29). Patients suffered from hyperlysinemia, nonketotic hyperglycinemia, disorders of branchedchain amino acid metabolism, and argininosuccinase deficiency have been detected by this procedure. 3. This instrument is a low-pressure gradient system. A high-pressure gradient system can be used for the analysis. We recommend that the buffers in a mobile phase be premixed with organic solvents (5% to 10% buffer in organic solvent and 5% to 10% organic solvent in buffer solution), and thorough degassed be performed when a high-pressure gradient system is being used.

134

Liu

4. The contribution to dispersion by the instrument itself must be minimized. Two detectors should be physically close each other. The internal diameter of the capillaries should be as small as possible. 5. When contamination was detected (baseline drift during the gradient elution) or separation deteriorated, and the change of guard column could not correct the problem, the first column (next to the guard column) was discarded, the second one moved to the first position, and a new column was added at second position. Under a normal situation, the guard or analytical column can be used for analysis up to 100 or 500 plasma samples, respectively. 6. Other individual amino acid (such as argininosuccinic acid) can be added to this amino acid standard mixture for peak identification and quantitation. 7. The purity of the reagent was important for extending detection limit. A previously unused and aged reagent (> 24-h old) was necessary for a higher sensitivity level detection (21). The sensitivity depended mainly on the ability to eliminate or subtract background levels and reduce the interfering substances present in solvents and reagents. 8. It is recommended that the concentration of the buffer solutions in the mobile phase be less than 0.1 M to avoid the potential precipitation during the gradient elution. 9. A siliconized tube is recommended to prevent blood platelets becoming activated with the resultant release of Tau and phosphoethanolamine (PEA). Hemolysis should also be prevented because it may lead to false increases in the concentrations of Asp, Glu, Tau, and PEA (4). 10. The time of blood withdrawal and the relation to dietary intake may be another reason that may cause varying results between laboratories. If it is possible, the blood sample should be collected between 8 and 9 AM before breakfast. 11. Many analytical factors could result in variation in the plasma amino acid profile. These include: (1) sample collection (2) centrifugation and ultrafiltration (3) storage conditions (4) contamination by platelets and leukocytes or hemolysis (5) shift in baseline during chromatographic analysis, and (6) interference or overlap of one or more amino acids. The physiological factors such as circadian rhythm, protein intake, pregnancy, nutritional status of the subject, physical activity before blood collecting, menstrual cycle, sex, and age differences and medication that interferes with the analysis may also have influence on plasma amino acids. All of these factors should be controlled to assess with certainty the clinical data under various pathological conditions. 12. Although most of the amino acids remained stable for up to 30 min at room temperature, delayed deprotenization may cause an increase of Asp, Glu, Gly, Ala, and Orn, whereas there is a decrease of Asn, Gln, His, and 3-MH (2). When deproteinization of physiological samples cannot be done immediately after centrifugation, it is recommended to store the specimens at temperature lower than – 18°C to prevent further hydrolysis of protein. In particular, the amino acids Asp and Glu increased markedly when the samples were not frozen immediately and not stored at –68°C or lower.


135

13. In dealing with physiological fluid samples, the protein should be completely removed from the sample prior to analysis. Direct injection onto RP columns are known to affect the retention behavior of the solutes and cause increase back pressure by irreversible adsorption of protein to the stationary phase. 14. Precaution should be taken when this method is used for some samples with higher protein concentration, such as blood cell lysates or tissue extracts. It has been reported that for the lipemia samples, the time required to filter sufficient sample was very variable and occasionally no filtrate could be obtained, presumably because the membrane pores got blocked. The alternative method is chemical precipitation by using various acids, bases, or organic solvents. Precaution should be taken to avoid hydrolysis, degradation, and oxidation of amino acids in selecting any such procedure for the deproteinization. The main factors to be considered are the amount of protein to be removed, the comparability or interference with the chromatographic analysis, the thermal and chemical stability, the solubility of the analytes, and finally, the cost, time, and labor. The use of 24– 30% SSA to deprotein 500-µL plasma samples (w/v) has been reported yield consistent results with high recoveries of most amino acids (2,4). Ethanol or acetonitrile (ACN) can also be used as precipitants (ACN or ethanol to plasma ratio 2–4:1, v/v) and has the advantage of resulting in full recovery of total Trp, whereas using SSA precipitation cannot completely recover this amino acid (4). 15. Several factors such as rotor, membrane, and the time of the ultrafiltration may have an influence on the final results of the ultrafiltration (11). These factors were investigated in our experiment. The recovery rates of amino acid in ultrafiltrates obtained by fixed-angle rotor were higher than that by swingingbucket rotor. This may be caused by the different polarization control between fixed-angle and swinging-bucket rotors. The use of a fixed-angle rotor provides polarization control. The angle counteracts the buildup of retained protein at the membrane surface, because this dense layer slides outward and accumulates at the edge of membrane. In a swinging-bucket rotor, the polarization layer is compacted over the entire membrane surface, restricting the passage of solute and solvents through the membrane. Our experiment suggests that the polarization occurring during the ultrafiltration may be one of the important factors that can influence the recovery rate of amino acids. This might be the reason why some other authors could not achieve satisfactory results by ultrafiltration, because it was difficult to control the polarization when ultrafiltration was performed under nitrogen pressure, or a syringe was used as the driving force for the ultrafiltration. 16. The recovery from ultrafiltration and its reproducibility of standard amino acids were tested by 10 aliquots of amino acid standards: 5 aliquots were analyzed by HPLC before ultrafiltration, another 5 aliquots were analyzed after ultrafiltration. The results showed that the recovery of all of the standard amino acids was excellent, ranging from 95–102% (11). The accuracy of measurement was tested by adding a known quantity of amino acid standards to a plasma sample, then the sample was ultrafiltrated and derivatized for analysis. The analytic recovery rate

136

Liu

Table 5 Recovery of Amino Acid Standard Added to Plasma Amino Acida

Plasma alone (µmol/L)

Plasma+250 mmol/L (expected)

Actual

Recovery Rate %

Ops Asp Glu GSH Asn Ser Gln Gly Thr His Cys Cit Tau Ala Arg Tyr AABA Met Val N-Val Trp Phe Ile Orn Leu Lys Hyp Pro

32.65 18.98 28.76 — 35.89 128.27 538.74 219.74 132.14 31.71 64.12 18.36 38.85 376.32 103.50 62.38 29.09 19.60 170.04 — 14.26 49.82 60.21 50.18 100.92 137.24 24.49 180.86

282.65 268.98 278.76 250.00 285.89 378.27 788.74 469.74 382.14 281.71 314.12 268.36 288.85 626.32 353.50 312.38 279.09 269.60 420.04 250.00 264.26 299.82 310.21 300.18 350.92 387.24 274.49 430.85

271.87 273.34 290.15 80.73 284.52 401.06 757.19 465.04 384.32 290.16 307.83 287.14 291.73 613.79 352.22 303.01 287.46 258.82 407.73 252.36 214.05 297.42 313.08 285.17 349.15 385.69 277.23 439.46

97 101 104 32 100 106 96 100 100 103 98 107 101 98 100 97 103 96 97 101 81 99 100 95 100 99 101 102

aAmino

acid abbreviations as in Table 3. Reprinted from ref. 8, p. 3331 by courtesy of Marcel Dekker, Inc.

of each amino acid was calculated after HPLC quantitation (Table 5). It should be mentioned that when standard amino acids were added to the plasma, Trp and GSH (a peptide) showed poor recovery (81% and 32%, respectively), whereas all of the other amino acids remained at a similar recovery. The reason was unknown. Changing the membrane in the ultrafiltration system to a membrane with higher molecular weight cutoff might be helpful for the improvement of the recovery rate of Trp and GSH.


17.

18.

19.

20.

21.

137

The multiplier in calibration form was set at 1.19 for the quantitation of plasma tryptophan. No effort has been made for the determination of GSH in plasma. Although the stability of OPA/3-MPA derivative is better than OPA/2-ME derivative, the OPA/3-MPA derivative of various amino acids does not have the same stability (8). When the waiting time was increased beyond 2.5 min, the stability of various amino acids was different. The UV absorbency of GSH and OPS increased, whereas asparagine, glutamine, glycine, histidine, taurine, ornithine, and lysine decreased. Others were unchanged. Therefore, the precise control of reaction time is still important in the derivatization using OPA/3-MPA as reagents. The results from our experiment indicated that OPA/3-MPA derivatives were relative stable during the waiting time from 0.5 to 5 min (8). When analyzing physiological samples with interested amino acids covering a broad retention range, it is necessary to modify the elution conditions during the analysis to optimize the separation. The combinations of the organic solvents and buffer solution; the pH and ion concentration of the buffer in the mobile phase; the patterns of gradient elution, and the different column and column temperature should be investigated to optimize the separation. Comparison of buffer pH from 6.4 to 7.2 was made in our experiment with all other conditions held constant (8). The pH at 6.8 appeared to give optimal separation. The pH of the buffer in the mobile phase should be close to neutral. A lower pH of the buffer would result in weaker UV absorbency and fluorescent intensity. Maintaining a stable column temperature is critical to the reproducibility of retention time, in addition to improving the separation of amino acids. When the column temperature was kept at 40°C, the coefficients of variation for retention time ranged from 0.02% to 0.97%, with a mean of 0.26%+0.21 SD. Column temperature was also tested for its effect on separation. When column temperature was maintained at 30°C, the separation of amino acids at first half of the chromatogram was better than that at temperature 35°C or 40°C, but the separation of second half was not satisfactory. Increasing the temperature to 40°C gave better separation for amino acids eluted later and the overall results were the best, together with a lower-column pressure. We therefore set the column temperature at 40°C for the analysis (8). The OPA/3-MPA and FMOC-Cl derivatives were detected in our experiment by a photo-diode array detector at three different sample wavelengths. The intensity of the signals, the baseline noises, and interference were different at various wavelength settings. At sample wavelengths of 338 nm and 266 nm, we compared two reference wavelengths at 390 nm and 550 nm. The use of 550 nm as reference wavelength produced more stable baseline than that at 390 nm. Therefore, 550 nm was used as reference wavelength for the photo-diode array detector. Under the chromatographic conditions we had chosen, there was no interfering peak on either UV or fluorescence signals during a blank solvent gradient elution. When HPLC water was used as a blank sample, there was no peak interfering with amino acid peak at UV sample wavelength 338 nm, and fluorescence signals. But at UV 266 nm, and especially at 230 nm, several peaks were

138

Liu

observed that were large enough and close enough to interfere with about 10 amino acid peaks. Although detection at UV 230 nm was more sensitive than at 338 nm, the interference could seriously compromise the quantitation. For this reason, UV 338 nm is recommended as the main sample wavelength for the quantitation of primary amino acids whereas wavelength 266 nm is recommended for the FMOCCl derivatives of secondary amino acids. Extremely high sensitivity is not required for the analysis of plasma amino acids, because the limitation is not sample size, rather the amount of plasma required for deproteinization procedure (such as ultrafiltration). 22. For the fluorescence detector we used to detect OPA and FMOC-Cl derivatives, both 230 nm and 340 nm were evaluated as Ex for OPA primary amino acid derivatives. The results showed that the detection of OPA primary amino acid derivatives at Ex 230 nm gave a response over seven times stronger than that at 340 nm, with an acceptable baseline. For the detection of FMOC-Cl secondary amino acid derivatives, three excitation wavelengths of 254, 260, and 266 nm were compared. The results showed 260 nm had the strongest signal. Therefore, we set Ex 230 nm and Em 450 nm during the first 20 min for the measurement of OPA derivatives of primary amino acids, then switched to Ex 266 nm and Em 315 nm for the determination of FMOC-Cl derivatives of secondary amino acids. The hydroxyproline, sarcosine, and proline were all eluted after 20 min in our chromatographic system. 23. Analysis of physiological samples with different amino acid concentrations is required during the practical applications. The linear relationship of amino acid concentrations in the ultrafiltrates was investigated. The linearity of response was estimated by injecting derivatized amino acids with different concentrations and constructing regression equations for UV and fluorescence response-concentration curves. The linear relationship between the concentration and peak areas of each standard amino acid was determined by analyzing the standard amino acid mixture at concentrations ranging from 31.25 to 500 µmol/L (n = 5, by serial dilution). These concentrations cover the normal range of most plasma amino acids. For plasma, the original plasma sample and plasma samples diluted by HPLC water to 75%, 50%, and 25% (n = 4) of the plasma were analyzed. The linear regression analysis showed satisfactory coefficients of correlation (>0.99) between the concentration and peak areas of each amino acid from both UV and fluorescent signals and in both standard amino acids and plasma samples (11).

References 1. Deyl, Z., Hyanek, J., and Horakova, M. (1986) Profiling of amino acids in body fluids and tissues by means of liquid chromatography. J. Chromatog. 379, 177– 250. 2. Qureshi, G. A. and Qureshi, A. R. (1989) Determination of free amino acids in biological samples: problems of quantitation. J. Chromatog. 491, 281–289.


139

3. Williams, A. P. (1987) General problems associated with the analysis of amino acids by automated ion-exchange chromatography. J. Chromatog. 373, 175–190. 4. Fekkes, D. (1996) State-of-the-art of high-performance liquid chromatographic analysis of amino acids in physiological samples. J. Chromatog. B. Biomed. Appl. 682, 3–22. 5. Uhe, A. M., Collier, G. R., McLennan, E. A., Tucker, D. J and O’Dea, K. (1991) Quantitation of tryptophan and other plasma amino acids by automated pre-column o-phthaldialdehyde derivatization high-performance liquid chromatography: improved sample preparation. J. Chromatog. 564, 81–91. 6. Sarwar, G. and Botting, H. G. (1993) Evaluation of liquid chromatographic analysis of nutritionally important amino acids in food and physiological samples. J. Chromatog. 615, 1–22. 7. Furst, P., Pollack, L., Graser, T. A., Godel, H., and Stehle, P. (1990) Appraisal of four pre-column derivatization methods for the high-performance liquid chromatographic determination of free amino acids in biological materials. J. Chromatog. 499, 557–569. 8. Worthen, H. G. and Liu, H. (1992) Automatic pre-column derivatization and reversed-phase high performance liquid chromatography of primary and secondary amino acids in plasma with photo-diode array and fluorescence detection. J. Liq. Chromatogr. 15, 3323–3341. 9. Carducci, C., Birarelli, M., Leuzzi, V., Santagata, G., Serafini, P., and Antonozzi, I. (1996) Automated method for the measurement of amino acids in urine by highperformance liquid chromatography. J. Chromatog. A. 729, 173–180. 10. Fekkes, D., van Dalen, A., Edelman, M., and Voskuilen, A. (1995) Validation of the determination of amino acids in plasma by high- performance liquid chromatography using automated pre-column derivatization with o-phthaldialdehyde. J. Chromatog. B. Biomed. Appl. 669, 177–186. 11. Liu, H. and Worthen, H. G. (1992) Measurement of free amino acid levels in ultrafiltrates of blood plasma by high-performance liquid chromatography with automatic pre-column derivatization. J. Chromatog. 579, 215–224. 12. Terrlink, T., van, L. P., and Houdijk, A. (1994) Plasma amino acids determined by liquid chromatography within 17 minutes. Clin. Chem. 40, 245–249. 13. van Eijk, E. H., Rooyakkers, D. R., and Deutz, N. E. (1993) Rapid routine determination of amino acids in plasma by high- performance liquid chromatography with a 2–3 microns Spherisorb ODS II column. J. Chromatog. 620, 143–148. 14. Georgi, G., Pietsch, C., and Sawatzki, G. (1993) High-performance liquid chromatographic determination of amino acids in protein hydrolysates and in plasma using automated pre-column derivatization with o-phthaldialdehyde/2-mercaptoethanol. J. Chromatog. 613, 35–42. 15. Liu, H., Liu, Z. M., Zhu, W. N., Li, Y. H., Wang, S. D., and Jiang, X. S. (1984) Analysis of free amino acids in blood plasma by reversed phase high performance liquid chromatography with gradient elution and fluorescence detection. Chinese J. Chromatog. 1, 83–87.

140

Liu

16. Ersser, R. S. and Davey, J. F. (1991) Liquid chromatographic analysis of amino acids in physiological fluids: recent advances. Med. Lab. Sci. 48, 59–71. 17. Jones, B. N. and Gilligan, J. P. (1983) o-Phthaldialdehyde precolumn derivatization and reversed-phase high- performance liquid chromatography of polypeptide hydrolysates and physiological fluids. J. Chromatog. 266, 471–482. 18. Schuster, R. (1988) Determination of amino acids in biological, pharmaceutical, plant and food samples by automated precolumn derivatization and high-performance liquid chromatography. J. Chromatog. 431, 271–284. 19. Turnell, D. C. and Cooper, J. D. (1982) Rapid assay for amino acids in serum or urine by pre-column derivatization and reversed-phase liquid chromatography. Clin. Chem. 28, 527–531. 20. Ogden, G. and Foldi, P. (1984) Amino acid analysis: An overview of current methods. LC-GC 5, 28–40. 21. Lindroth, P. and Mopper, K. (1979) High performance liquid chromatographic determination of subpicomole amounts of amino acids by precolumn fluorescence derivatization with o-Phthaldialdehyde. Anal. Chem. 51, 1667–1674. 22. Einarsson, S., Josefsson, B., and Lagerkvist, S. (1983) Determination of amino acids with 9-fluorenylmethyl chloroformate and reversed-phase high-performance liquid chromatography. J. Chromatog. 282, 609–618. 23. Einarsson, S. (1985) Selective determination of secondary amino acids using precolumn derivatization with 9-fluorenylmethylchloroformate and reversed-phase high-performance liquid chromatography. J. Chromatog. 348, 213–220. 24. Betner, I. and Foldi, P. (1988) The FMOC-ADAM approach to amino acid analysis. LC-GC 6, 832–840. 25. Godel, H., Graser, T., Foldi, P., Pfaender, P., and Furst, P. (1984) Measurement of free amino acids in human biological fluids by high-performance liquid chromatography. J. Chromatog. 297, 49–61. 26. Blundell, G. and Brydon, W. G. (1987) High performance liquid chromatography of plasma aminoacids using orthophthalaldehyde derivatisation. Clin. Chim. Acta 170, 79–83. 27. Feste, A. S. (1992) Reversed-phase chromatography of phenylthiocarbamyl amino acid derivatives of physiological amino acids: an evaluation and a comparison with analysis by ion-exchange chromatography. J. Chromatog. 574, 23–34. 28. Armstrong, M. D. and Stave, U. (1973) A study of plasma free amino acid levels. II. Normal values for children and adults. Metabolism 22, 561–569. 29. Armstrong, M. D. and Stave, U. (1973) A study of plasma free amino acid levels. III. Variation during growth and aging. Metabolism 22, 571–578.

HPLC Analysis of BTC, BZTC-Amino Acids

141

11 Determination of Amino Acids in Foods by Reversed-Phase High-Performance Liquid Chromatography with New Precolumn Derivatives, Butylthiocarbamyl, and Benzylthiocarbamyl Derivatives Compared to the Phenylthiocarbamyl Derivative and Ion Exchange Chromatography Kang-Lyung Woo 1. Introduction Amino acid analysis with reverse-phase high-performance liquid chromatography (RP-HPLC) and ultraviolet (UV) detection following precolumn derivatization is popular owing to the greater versatility at the instrument, sensitivity and speed of analysis compared to specialized ion–exchange amino acid analyzers. Phenylthiocarbamyl (PTC) amino acid derivative is a precolumn derivatization method that has been widely used for analysis of amino acid by RP-HPLC (1–8). This method is an excellent method for the derivatization of secondary amino acids, proline, and hydroxyproline (2–4,7). However, its disadvantages are that it requires a high-vacuum system and it takes a long time to remove the byproducts produced in the process of derivatization and the excess reagent in order to avoid interfering peaks. A need for simple, sensitive, stable, and more volatile precolumn derivatization reagent for analysis of amino acid with RP-HPLC and UV-detection still remains, although there have been developments of many reagents for precolumn derivatization, because these derivatives have some faults. The widely used precolumn derivatives, except PTC-amino acids, in RP-HPLC are o-phthalaldehydes (OPA) (9,10), dansyl (11,12), dabsyl (13,14), and 4-nitrophenylthiocarbamyl (NPTC) (7,15) derivatives. With the OPA derivatives, secondary amino acids, proline, and hydroxyproline, were not detected From: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

141

142

Woo

because OPA does not react with secondary amines in the absence of oxidizing agents. Moreover, as OPA derivatives are unstable, complete automation of the precolumn reaction with accurate control of the reaction time is essential for acceptable reproducibility (1). Dansyl derivatives are formed in the dark and are unstable toward prolonged reaction times, solvents, and exposure to light and interfering peaks arise because of the byproducts during the derivatization process (12). In the dabsyl derivatives, it was reported that the sulphonamide bond in the derivative was very stable. The limitation of the dabsyl derivatization method is that the presence of an excess amount of urea, salt, phosphate, or ammonium hydrogen carbonate will change the pH of the reaction buffer and interfere with derivatization (13,14). Derivatization with 4-nitrophenylisothiocyanate (NPITC) forms the stable nitrophenylthiocarbamyl (NPTC) derivatives that are suitable for analysis by RP-HPLC and UV-detection at 254 nm or 340 nm (15). The disadvantage of NPITC is that the excess reagent cannot be easily removed under a highvacuum system. Extraction with toluene can remove the excess reagent (7). A more volatile reagent for precolumn derivatization compared to other reagents developed to recent year, butylisothiocyanate (BITC), an aliphatic compound, was successfully adopted for derivatization to butylthiocarbamyl (BTC) derivatives of 22 protein standard amino acids. The BTC-amino acids were successfully analyzed via C18 RP-HPLC and UV-detector at 250 nm (16). The BITC reagent was also successfully adapted for the analysis of amino acids in foods (17). The advantages of BITC were high volatility and the ability of separate derivatization on the cysteine and cystine, which had not been found with PTCderivatives. The high volatility of this reagent substantially reduced the analysis time because the excess reagent and byproducts produced during the reaction could be easily removed. The BTC-derivatives of the secondary amino acids, proline, and hydroxyproline were also detected with high sensitivity. But asparagine and serine peaks overlapped completely and the stability of BTC-derivatives at room temperature was estimated to be only approx 8 h. Benzylisothiocyanate (BZITC), the analog of phenylisothiocyanate (except NPITC), was successfully derivatized to benzylthiocarbamyl (BZTC) derivatives on all of the 22 protein amino acids and the derivatives were completely separated on a reversed-phase Nova-Pak C18 column (18). The BZITC reagent was less volatile compared to BITC but the volatility was similar to PITC. The advantages of BZTC-derivatives compared to PTC-derivatives was superior resolution on reversed-phase column and superior reproducibility to PTC-derivatives using the same experimental condition.


143

In this chapter, more detail experimental methods, results, and discussions on the BTC and BZTC-derivatives developed at the most recent years will be described. 2. Materials 2.1. Equipment 1. A water aspirator for vacuum (see Note 1). 2. Spectra-Physics 8800 ternary solvent delivery system with solvent stabilization and degassing system. 3. Spectra 200 programmable wavelength UV detector. 4. Nova-Pak C18 (300 × 3.9 id, 4 µm dimethyloctadecylsilyl-bonded amorphous silca, Waters). 5. Eppendorf CH-30 column heater.

2.2. Reagents 1. BITC, BZITC, and PITC are obtained from Aldrich (Milwaukee, WI) and were stored at 0–5°C (see Note 2). 2. Bovine serum albumin (BSA), standard amino acids, and norleucine are obtained from Sigma (St. Louis, MO). Standard amino acids were stored at room temperature and BSA was stored at 2–8°C. 3. HPLC-grade acetonitrile, methanol, and tetrahydrofuran from Merk (Darmstadt, Germany) were stored at room temperature. 4. All other reagents were of analytical grade. 5. Food samples, whole egg, and soybean purchased from a commercial market.

2.3. Solutions 1. Standard amino acid solution: A mixture solution of standard amino acids, except glutamine, cysteine, and cystine was prepared at a concentration of 2.5 µmol/mL of 0.01 M HCl. Standard solutions of glutamine and cysteine were prepared with water (see Note 3). Cystine was prepared at a concentration of 0.5 µmol/mL of 0.01 M HCl because of the solubility. 2. Preparation of coupling buffer solution: Coupling buffer (acetonitrile-methanoltriethylamine [10:5:2]) containing L-norleucine (2.5 µmol/mL) as an internal standard (see Note 4) was stored at 0–5°C and prepared again after 1 mo.

3. Method 3.1. Hydrolysis of BSA and Food Protein Samples (Figs. 1 and 2) 1. Place BSA (4 mg) and grinded soybean or homogenized whole egg food sample (0.2 g) in a 5-mL and 25-mL test tube, respectively, with an open-hole screw cap with a septum (Fig. 3). 2. Add 0.5 mL and 15 mL of 6 M HCl containing 0.1% phenol into the 5-mL and 25mL test tubes containing the BSA and food samples, respectively.

144

Woo

Fig. 1. Flow chart for hydrolysis of food samples.

Fig. 2. Flow chart for hydrolysis of BSA.


145

Fig. 3. Test tube for hydrolysis of food protein samples.

Fig. 4. The device for supply of the dried N2 gas. 3. After tightly capping, pierce the septum with two stainless steel injection needles. 4. Connect one needle, immersed in the sample solution, to a dried nitrogen supply (see Note 5, Fig. 4). Connect the other needle, not immersed, to the vacuum pump (Fig. 5). 5. Evacuate the test tubes with a vacuum pump for 5 min and simultaneously flush with nitrogen gas.

146

Woo

Fig. 5. The diagram of device for nitrogen saturation into the sample.

Fig. 6. Flow chart for clean up with cation–exchange column. 6. Remove the needle connected to the vacuum pump before removing the needle connected to the nitrogen. 7. Carefully remove the cap with the holes and change for a cap without holes (see Note 6), because small holes may be formed by the high pressure. 8. Carry out hydrolysis at 145°C for 4 h. 9. Dry the hydrolysate of BSA with nitrogen at 50°C. Redissolve with 5 mL of 0.01 M HCl (hydrolysate solution). This is now ready for derivatization.


147

10. Filter the hydrolysates of soybean meal and whole egg, then dry with a rotary evaporator. Redissolve and adjusted the volume to 50 mL with 0.01 M HCl (hydrolysate solution). 11. To clean up the hydrolysate solutions of soybean meal and whole egg use cation– exchange chromatography as follows (Fig. 6.). 12. Pass 5 mL of hydrolysate solution through a 100 × 13-mm id cation–exchange column (Dowex 5 × 8) at the rate of 6 drops/min, to retain the amino acids on the cation–exchange resin (see Note 7). 13. Wash the column several times with 20 mL of H2O (see Note 8). 14. Elute the retained amino acids on the cation–exchange column with 40 mL of 4 M ammonia solution with the rate of 6 drops/min. 15. Dry the eluted solutions in a rotary evaporator at 50°C. Redissolve in 0.01 M HCl (hydrolysate solution) and adjust the volumes to 50 mL. These samples are now ready for derivatization (see Note 9).

3.2. Derivatization (Fig. 7) 1. Place 20 µL of the mixture solution of standard amino acids, 50 µL of the standard solution of cystine, and sample hydrolysates (BSA; 100 µL, soybean; 500 µL, whole egg; 500 µL) into separate 2-mL conical vials with an open-hole screw cap and a septum (Fig. 8). 2. Dry the solutions completely with nitrogen gas at 50°C. 3. Add an appropriate amount of acetonitrile to each vial and dried again (see Note 10). 4. Redissolve the residues in 50 µL of coupling buffer. 5. Add of 3 mL of BITC, BZITC, and PITC to each of the dissolved solutions for BTC, BZTC, and PTC derivatives, respectively. 6. After tightly capping the vials with open-hole screw-caps, the derivatizations are carried out at 40°C for 30 min for BTC and PTC derivatives and at 50°C for 30 min for BZTC derivatives. 7. After derivatizations, use two stainless steel injection needles to pierce through the septum into the vials. Connect one needle with the nitrogen supply and the other with the vacuum pump (Fig. 9.). 8. Infuse nitrogen into the vials and simultaneously evacuate with the vacuum pump to complete dryness at room temperature for approx 10 min for BTC derivative and for approx 40 min for BZTC and PTC derivatives. 9. Inject 100 µL of acetonitrile into the vials with a microinjection syringe and redry the contents for 5 min for BTC-derivative and for 30 min for BZTC and PTC derivatives (see Note 11). 10. Dissolve the residue of BTC-derivatives in 1 mL of 0.02 M ammonium acetate. 11. Dissolve the residue of BZTC and PTC-derivatives in 1 mL of 0.02 M NaH2PO4 containing 5% methanol and 1.5% tetrahydrofuran (pH 6.8, adjusted with phosphoric acid). 12. Filter the dissolved solutions through a 0.25-µm membrane filter. 13. Inject 10-µL aliquots of the filtrates onto the respective HPLC system.

148

Woo

Fig. 7. Flow chart for derivatization.


149

Fig. 8. 2-mL conical vial for derivatization.

Fig. 9. The system drying several vials simultaneously.

3.3. Chromatography HPLC conditions for the analysis were elucidated with Table 1.

3.4. Sensitivities of BTC, BZTC, and PTC-Derivatives The sensitivities of BTC, BZTC, and PTC-derivatives are detected at 0.05 AUFS, which is the limit that gives a stable baseline with the smallest quantity. It is possible to determine a linear relationship with the quantitative analysis (see Notes 21 and 22).

150

Woo

Table 1 HPLC Conditions for the Analysis of BTC, BZTC, and PTC-Amino Acids Derivatives HPLC system Detector Wavelength for detection Column

Solvent system

Spectra-Physics 8800 ternary solvent delivery system Solvent stabilization and degassing system with a blanket of helium Spectra 200 programmable wavelength UV detector 240 nm for BTC-derivatives (see Note 12) 246 nm for BZTC-derivatives (see Note 13) 254 nm for PTC-derivatives (see Note 13) Nova-Pak C18 (300 × 3.9 id, 4 µm dimethyloctadecylsilyl-bonded amorphous silca, Waters). For all derivatives column-temperature; 40°C with Eppendorf CH-30 column heater For BTC-derivatives (see Notes 14–17): A solution; 0.05 M ammonium acetate (pH 6.7 adjusted with phosphoric acid) B solution; 0.02 M sodium phosphate dibasic solution containing 5% methanol and 1.5% tetrahydrofuran-acetonitrile (50:50) C solution; acetonitrile-water (70:30) Solvent gradient A B C Flow rate 0.0 min 100% 0% 0% 1 mL/min 5.0 min 85 15 0 " 14.0 min 70 20 10 " 20.0 min 60 20 20 " 25.0 min 30 20 50 " 30.0 min 10 20 70 " For BZTC-derivatives (see Notes 14–17) and PTC-derivatives (see Note 14–17) A solvent; 0.02 M NaH2PO4 containing 5% methanol and 1.5% tetrahydrofuran (pH 6.8 adjusted with phosphoric acid) B solvent; A solvent-acetonitrile (50:50) C solvent; acetonitrile-water (70:30) Solvent gradient A B C Flow rate 0.0 min 100% 0% 0% 1.2 mL/min 15.0 min 76 20 4 " 20.0 min 70 20 10 " 30.0 min 50 30 20 " 40.0 min 30 35 35 " After this gradient program, a washing step for 20 min with solvent C substantially protected the column damage. Above two types of solvent gradients were adapted to the samples amino acid derivatives as well as standard amino acid derivatives (see Note 18–20).


151

3.5. Stabilities of the BTC-Derivatives and BZTC-Derivatives The variations in the peak-area responses of BTC and BZTC-derivatives with storage time at room temperature can be determined for the stabilities of the BTC and BZTC-derivatives (see Notes 23 and 26).

3.6. Statistical Analysis 1. To determine the reproducibilities, all experiments should be repeated more than three times. 2. Relative standard deviations (RSDs) on the relative molar response (RMR) can be calculated for comparision of precisions on the derivatives. Linearity of calibration graphs on the appropriate ranges are also detected by the determination of statistical significance of correlation coefficient (γ) of calibration graphs (see Notes 27 and 33). 3. The reproducibility and accuracy on the BSA and food samples can be compared to ion–exchange chromatography and the data of the other literatures (see Notes 34 and 38).

4. Notes 1. If you use a vacuum pump that is not using water, you must install a device to absorb the evaporating excess reagents and byproducts produced during derivatization because the vacuum pump is rendered useless by these evaporating materials. 2. These reagents seem to be stable for several years unopened, but when the caps were opened several times, the reagents must be used in about 6 mo even if they are stored at –20°C. All of these reagents are very harmful and toxic. 3. Standard solutions of glutamine and cysteine were prepared with water because of the conversion to pyroglutamic acid and cystine, respectively, on prolonged storage of these amino acids in HCl solution (19). 4. For BZTC-derivatives of sample internal standard (L-norleucine) should not be used. 5. For supply of the dried nitrogen, connect the tube for absorption of any moisture as shown in Fig. 2. This Na2SO4 anhydrous tube must be periodically changed every 1 or 2 mo. 6. When the cap with a hole is changed to the cap without a hole, you have to be carefully that the septum is not removed. 7. Be careful that the sample, washing, and eluting solutions do not drop under the level of cation–exchange resin in the column. 8. The rate of drops during the washing is not important. 9. For the analysis with ion–exchange chromatography and ninhydrin derivatization method, dry the clean-up sample solutions in a rotary evaporator. Redissolve in 0.2 M sodium citrate buffer (pH 2.2) and inject into the automation amino acid analyzer (LKB 4150 Alpha, Ultrapac-11 cation–exchange column). 10. By the synergy effect of acetonitrile evaporating, the residues of water are completely dried.

152

Woo

11. The chemical reactions of BTC and BZTC-derivatives are as follows:

12. UV-spectra of the BTC-amino acid mixture and BITC, the coupling reagent, are shown in Fig. 10. The λmax of BTC-amino acids was about 234 nm, but the most efficient wavelength was 250 nm, which avoided the absorption spectra of the impurities and the electrolyte, ammonium acetate in the solvent (16). 13. The UV spectra of the BZTC-amino acids mixture and the PTC-amino acid mixture dissolved in 0.02 M NaH2PO4 are shown in Fig. 11. The wavelengths giving strong absorbance were 220 nm and 238 nm in the BZTC derivative and 215 and 270 nm in the PTC derivative, but the most efficient wavelengths were 246 nm in the BZTC derivative and 254 nm in the PTC derivative, which avoided interference by the absorption spectra of the impurities and electrolyte and showed a stable baseline (18). 14. The standard amino acid chromatograms of BTC and BZTC-derivatives compared to PTC-derivatives separated on the Nova-Pak C18 column are shown in Fig. 12. All of the 22 standard amino acids were derivatized with BTC, BZTC, and PTC-derivatives and resolved on C18 reversed column. 15. In the BTC-derivatives, asparagine and serine completely overapped but BTCcysteine and cystine were individually eluted even though the cystine peak was resolved with a tailing peak. In the PTC-derivatives, cystine and cysteine peaks appeared at the same position, from which it could be assumed that cysteine might be completely converted into cystine during derivertization because of the fact that cysteine can be oxidized (20). Other articles have also reported that PTCcysteine and cystine were eluted at the same position (2,4,7).


153

Fig. 10. UV spectra (x-axis in nm) of BTC-amino acid mixture and the coupling reagent BITC. Absorbance of solvents at 250 nm = 0.

Fig. 11. UV spectra of the BZTC-amino acid mixture and the PTC-amino acid mixture. Solvent, 0.05 M NaH2PO4. Absorbance of solvent at 254 nm = 0.

154

Woo

Fig. 12. Chromatogram of standard protein amino acid derivatives resolved on a Nova-Pak (30 cm × 3.9 mm) C18 column. I.S. = norleucine, injected amount 0.625 nmol. (A) BTC-amino acid; (B) BZTC-amino acid; (C) PTC-amino acid. Cyt = cystine; Cys = cysteine.


155

Fig. 13. Absorbance plot of BZTC-cysteine derivative at 238 nm showing maximum absorption of the mole fraction of BZITC in the derivatizing solution for the reaction between benzylisothiocyanate and cysteine. 16. Unlike BTC and PTC-derivatives, in the BZTC-derivatives all of 22 standard amino acids were nearly completely separated. Also, in the BZTC-derivatives as in the BTC-derivatives, the cysteine and cystine peaks were markedly separated. 17. Unlike the PTC-derivatives, in the BZTC and BTC-derivatives, the -SH moiety of cysteine would be converted to the thiocyanate derivatives by BZITC and BITC. To find out whether 1 mol of cysteine reacts with 2 mol of BZITC, absorbances of the BZTC-derivatives on the mole fraction of BZITC in the solution for derivertization between BZITC and cysteine were determined (Fig. 13). The determined wavelength was 238 nm, which gave maximum absorption for the BZTC-derivatives. Maximum absorbance appeared at 0.667 of the mole fraction, which meant that the mole ratio of the reaction between cysteine and BZITC was 1:2. So we could draw the conclusion that the -SH group, as well as the amino group of cysteine was derivatized. Intelligence due to the fact that -SH moiety of cysteine was readily converted to the thiocyanate derivative with 2-nitro-5thiocyanobezoic acid (21) could support this conclusion. BZTC and BTC-cysteine might be eluted last because of the long less-polar side chain by the derivatization of –SH group. 18. Chromatograms of BTC, BZTC, and PTC-amino acid derivatives of soybean meal hyrolysate are shown in Fig. 14. In the BTC and PTC-derivatives we could detect the same kinds of amino acids of 16, but in the BZTC-derivatives cystine and cysteine in addition to above 16 amino acids were detected. We think that the sensitivity of BZTC-derivatives are more superior to the BTC and PTC-derivatives. 19. Chromatograms of BTC and PTC-amino acid derivatives on the whole egg hydrolysate are shown in Fig. 15. Because unfortunately, we did not detect on the BZTC-amino acid derivatives of the whole egg hydrolysate, it is really regretable that we could not compare to BTC and PTC-derivatives. The same kinds of 17 amino acids in the BTC and PTC-derivatives were detected. As is shown in the chromatograms of food samples (Figs. 14 and 15), assuming that the some kinds

156

Woo

Fig. 14. Chromatogram of amino acids in soybean hydrolysate. (A) BTC derivatives; (B) BZTC derivatives; (C) PTC derivatives.


157

Fig. 15. Chromatogram of amino acids in whole egg hydrolyste. (A) BTC drivatives; (B) PTC derivatives. Cys = cysteine. of contaminants were contained there were few ghost peaks. This phenomenon indicates that BITC, BZITC, and PITC are very good reagents having superior selective reactivity with amino acids. 20. Chromatograms at BTC, BZTC, and PTC-derivatives of BSA hydrolysate are shown in Fig. 16. In the soybean meal, on the other hand, the BTC and PTCderivatives detected 17 amino acids, in the BZTC-derivatives 18 amino acids were detected with the additional detection of cystine. 21. Sensitivities of BTC, BZTC, and PTC-derivatives were about 3.9 pmol at 0.05 AUFS (Fig. 17). At levels lower than 3.9 pmol, several amino acids in the all of the three derivatives were not detected and not showed the linearity in the calibration graphs for the quantitative analysis. 22. There is a report that the sensitivity of the PTC-derivatives was 1 pmol in the detector at 0.005 AUFS (signal-to-noise ratio 5:1) (2). But in our experiment, at

158

Woo

Fig. 16. Chromatogram of amino acids in bovine serum albumin hydrolysate. (A) BTC derivatives; (B) BZTC derivatives; (C) PTC derivatives. Cys = cysteine.


159

Fig. 17. Chromatogram of standard protein amino acid derivatives showing the sensitivity. Injected amount 3.9 pmol. Range 0.05 AUFS. (A) BTC-amino acid; (B) BZTC-amino acid; (C) PTC-amino acid. Cyt = cystine; Cys = cysteine.

this level, because the contamination of amino acids by the reagents, instrument, environment, and solvents was serious and the baseline was very unstable, the adaptation of the practical analysis was impossible.

160

Woo

Fig. 18. Stability of BTC-amino acids at room temperature. Leu.

䊏

= Glu;

䊉

= Ile;

䉱=

Fig. 19. Stability of BZTC derivatives of hydroxyproline (䊏), proline (䉱), serine (䊉), cystine (䊐), alanine ( ), and threonine (䊊). Each point is the mean value of three measurements. 䉮

23. The stability of BTC-derivatives and BZTC-derivatives is shown in Figs. 18 and 19, respectively. 24. In the BTC-derivatives, up to 2 h of storage, the peak area responses increased for most of the derivatives, which indicated the derivatization reaction was continuing at room temperature. After 8 h of the storage, the derivative that showed


161

Table 2 RMR of BTC, BZTC, and PTC Derivatives of Standard Amino Acids Amino acid

BTC derivatives

BZTC derivatives

PTC derivatives

RMRa

RSD

RMRb

RSD

RMRb

RSD

Asp Glu Hyp Asn Gln Ser Gly His Pro Arg Thr Ala Val Tyr Ile Leu Cyt Phe Trp Lys Cys

0.50 0.48 0.07 0.66c 0.71 — 0.52 0.62 0.71 0.44 0.29 0.75 0.88 0.83 0.84 0.86 0.49 1.01 0.93 1.10 0.63

1.40 1.25 1.64 0.45 1.27 — 0.96 1.13 1.55 0.68 1.72 2.53 3.86 1.33 0.03 2.21 4.29 2.57 3.87 3.27 5.24

1.89 1.21 1.61 1.12 1.07 1.05 0.96 0.84 1.17 1.00 0.92 0.88 0.95 1.39 1.05 0.84 0.79 1.52 2.04 1.27 3.63

3.82 0.83 4.50 4.48 2.80 1.46 2.63 2.49 3.42 1.15 1.09 1.73 1.08 2.20 0.95 1.38 3.64 1.32 0.98 0.91 7.22

0.89 0.84 0.93 0.91 1.03 0.76 0.65 0.83 0.85 0.91 1.01 0.90 0.83 0.92 0.85 0.84 — 0.76 0.92 1.24 1.62d

2.48 1.83 1.62 2.37 5.89 2.02 1.68 0.87 7.37 3.02 16.6 6.64 4.68 2.64 3.76 6.86 10.5 2.46 1.64 10.8 —

RMR, relative molar response RSD, relative standard deviation (n = 3) aValues are relative to methionine. bValues are relative to norleucine. cAsn+Ser dCyt+Cys. Cyt = Cystine, Cys = cysteine.

the greatest decrease was cysteine (17.3%), the losses of the other derivatives after 8 h were in the range 0–6.9%. The derivatives that decreased to less than 5% after 14 h were glutamic acid, asparagine+serine, glutamine, threonine, tyrosine, proline, lysine, and tryptophan (16). 25. In the BZTC-derivatives, the peak area responses were decreased to less than 5% up to 120 h storage, but threonine, alanine, cystine, and serine were decreased to 18.9, 13.7, 15.1, and 15.6% after 4 h, respectively. Regarding the stability, we know that the optimum pH of the solvent dissolving the derivatives was very important. For the PTC-derivatives, the optimum pH of the solvent was 7.5. The loss of PTC-derivatives at this pH was in the range 0–10% after 10 h storage at room temperature (7).

162

Woo

Table 3 Amino Acid Compositions of Soybean Determined by the BTC, BZTC, and PTC-Derivatives Compared to the Ion–Exchange Chromatography (g/100 g dry matter) Amino BTC acid derivativesa

BZTC derivativesb

PTC derivativesa

PTC Ion-exchange derivativesb chromatographya

4.17 ± 0.24 6.96 ± 0.42 1.61 ± 0.12 1.59 ± 0.07 1.50 ± 0.06 3.38 ± 0.17 1.19 ± 0.04 1.72 ± 0.10 2.15 ± 0.12 1.20 ± 0.04 2.88 ± 0.24 0.42 ± 0.08 2.46 ± 0.14 3.38 ± 0.18 — 2.27 ± 0.16 3.03 ± 0.13 —

4.45 ± 0.25 7.41 ± 0.14 2.00 ± 0.16 1.56 ± 0.03 1.06 ± 0.08 3.61 ± 0.13 1.78 ± 0.11 2.00 ± 0.12 1.96 ± 0.20 1.20 ± 0.05 1.97 ± 0.10 0.73 ± 0.01 1.70 ± 0.05 2.95 ± 0.18 0.49 ± 0.12 2.06 ± 0.15 2.72 ± 0.10 0.17 ± 0.04

4.64 ± 0.28 7.25 ± 0.34 1.59 ± 0.02 1.77 ± 0.10 1.52 ± 0.04 2.28 ± 0.08 1.79 ± 0.17 1.89 ± 0.35 2.06 ± 0.08 1.29 ± 0.08 2.47 ± 0.10 0.54 ± 0.06 2.21 ± 0.06 3.03 ± 0.31 — 2.10 ± 0.10 3.15 ± 0.16 —

4.72 ± 0.63 7.76 ± 0.84 2.03 ± 0.43 2.09 ± 0.13 1.15 ± 0.04 2.65 ± 0.36 1.79 ± 0.05 1.89 ± 0.35 2.38 ± 0.18 1.26 ± 0.11 2.12 ± 0.10 0.50 ± 0.25 1.83 ± 0.16 3.10 ± 0.24 — 2.63 ± 0.18 3.73 ± 0.25 0.11 ± 0.01c

Asp Glu Ser Gly His Arg Thr Ala Pro Tyr Val Met Ile Leu Cyt Phe Lys Cys

4.25 ± 0.32 7.19 ± 0.12 1.65 ± 0.01 1.62 ± 0.05 1.32 ± 0.05 3.50 ± 0.09 1.24 ± 0.10 1.80 ± 0.12 2.20 ± 0.06 1.15 ± 0.05 3.01 ± 0.14 0.40 ± 0.03 2.25 ± 0.18 3.57 ± 0.21 — 2.19 ± 0.14 2.86 ± 0.19 —

All values are mean ± SD. an = 4. Data from ref. 17. bn = 5. Data from ref. 18. cCys+Cyt. Cyt = cystine, Cys = cysteine.

26. We assumed that the stability of the BZTC-derivatives was superior to those of the PTC-derivatives and BTC-derivatives, except for threonine, alanine, cystine, and serine. 27. The relative molar responses of BTC, BZTC, and PTC-derivatives were shown in Table 2. 28. In the BTC-derivatives, RMR values determined with relative to methionine. RSDs on the RMR were less than 5% except cysteine (5.24%) in the BTC-derivatives. In the BZTC and PTC-derivatives, RMR values were determined with relative to norleucine, the internal standard. 29. In the BZTC-derivatives, the R.S.D values of all derivatives except cysteine (7.22%) were less than 5%. However, for the PTC-derivatives of glutamine, proline, alanine, and leucine the RSDs exceeded 5% and for threonine, cystine+cysteine, and lysine the RSDs exceeded 10%. In the PTC derivatization,


163

Table 4 Amino Acid Compositions of Whole Egg Determined by the BTC and PTC-Derivatives Compared to the Ion–Exchange Chromatography (g/100 g dry matter) Amino acid

BTC derivatives

PTC derivatives

Ion-exchange chromatography

Asp Glu Ser Gly His Thr Ala Arg Pro Tyr Val Met Ile Leu Phe Lys Cys

4.20 ± 0.09 5.48 ± 0.05 3.50 ± 0.09 1.56 ± 0.09 1.90 ± 0.14 1.75 ± 0.13 2.71 ± 0.16 3.55 ± 0.10 1.95 ± 0.15 2.36 ± 0.62 3.64 ± 0.20 1.68 ± 0.11 3.42 ± 0.16 4.63 ± 0.21 2.94 ± 0.14 4.16 ± 0.20 1.35 ± 0.16

4.74 ± 0.07 5.85 ± 0.59 2.61 ± 0.42 1.54 ± 0.31 1.91 ± 0.15 2.15 ± 0.11 3.29 ± 0.19 3.43 ± 0.48 1.87 ± 0.24 1.96 ± 0.20 3.83 ± 0.36 1.58 ± 0.18 2.92 ± 0.28 4.32 ± 0.42 2.65 ± 0.22 4.30 ± 0.33 0.40 ± 0.10

4.23 ± 0.05 5.76 ± 0.07 3.54 ± 0.10 1.53 ± 0.02 1.79 ± 0.06 1.82 ± 0.15 2.70 ± 0.11 3.61 ± 0.16 1.98 ± 0.22 2.23 ± 0.30 3.59 ± 0.32 1.63 ± 0.06 3.26 ± 0.21 4.35 ± 0.09 2.88 ± 0.15 4.18 ± 0.30 1.41 ± 0.23

All values are mean ± SD, n = 4. Cys = cysteine.

the PITC reagent must be stored at –20°C under the inert gas to prevent breakdown (7) and a high-vacuum system (50–100 mtorr; 1 torr = 133.322 Pa) is needed to remove the excess reagent and byproducts that could interfere with the main peaks and to improve the reproducibility (3). In this chapter, we did not use the high-vacuum system, so this seems to be one of the reasons why the RSD values in the PTC-derivatives exceeded 5%. 30. We concluded that BTC and BZTC derivatization was less fastidious compared to PTC derivatization. 31. Calibration graphs for BTC-derivatives showed good linearity in the range 0.5– 2.5 nmol (not shown with Fig. or Table). The correlation coefficients (γ) of the calibration graphs for all of the derivatives were highly significant (p < 0.001), and the lowest value was γ = 0.926 for the cystine derivative, but it could be used for quantitative analysis. 32. Calibration graphs for all of the BZTC derivatives also showed good linearity in the measured range (0.125–5 nmol). The correlation coefficients (γ) of the calibration graphs were also highly significant (p < 0.001) and exceeded 0.99, except for cysteine (0.952) and glutamine (0.976).

Amino acid Asp Glu Ser Gly His Arg Thr Ala Pro Tyr Val Met Ile Leu Phe Lys Trp Cys Cyt Gln Asn Total

164

Table 5 Amino Acid Compositions of BSA Determined by BTC and PTC-Derivatives Method Compared to the Ion–Exchange Chromatography and the Analytical Results Reported in the Literature BTC derivativesa

BZTC derivativesc

PTC derivativesa

PTC derivativesa,c

Ion-exchange chromatographya

Automatic analyzerd

Sequencee

Sequencef

54.85 ± 3.01b 73.52 ± 5.04 24.60 ± 0.45 16.68 ± 0.28 22.39 ± 0.49 21.94 ± 0.22 30.35 ± 0.82 46.07 ± 0.94 26.85 ± 1.13 20.47 ± 0.55 38.61 ± 0.85 4.61 ± 0.28 15.25 ± 0.29 61.77 ± 0.26 29.20 ± 1.07 58.75 ± 1.97 — 26.82 ± 3.60 — — — 572.73

57.5 ± 4.76b,g 79.8 ± 5.54 24.4 ± 1.60 13.8 ± 1.42 16.4 ± 3.22 23.0 ± 2.55 32.4 ± 1.52 47.6 ± 2.24 28.3 ± 2.65 22.9 ± 1.56 38.2 ± 1.50 4.2 ± 0.25 15.9 ± 1.09 66.6 ± 2.02 28.7 ± 2.44 58.1 ± 2.73 — 12.4 ± 1.60 12.5 ± 1.63 — — 582.7

53.78 ± 4.75b 72.61 ± 0.94 26.22 ± 4.38 14.96 ± 0.50 20.85 ± 2.40 31.06 ± 1.51 31.16 ± 0.76 38.17 ± 0.27 29.27 ± 0.78 20.72 ± 0.60 36.87 ± 0.84 6.97 ± 0.48 18.67 ± 0.55 60.88 ± 0.65 29.26 ± 0.65 54.58 ± 4.65 — 9.32 ± 0.59 — — — 555.35

52.2 ± 0.42b,h 78.5 ± 2.92 22.3 ± 1.92 16.3 ± 2.03 18.4 ± 0.70 23.9 ± 1.80 32.6 ± 0.80 40.0 ± 0.28 32.7 ± 0.35 20.3 ± 2.12 40.2 ± 2.42 4.6 ± 0.56 16.9 ± 0.25 63.3 ± 3.68 30.3 ± 1.70 64.4 ± 2.42 — 25.8 ± 5.30g — — — 582.7

54.49 ± 3.51b 75.99 ± 4.14 25.84 ± 0.15 16.31 ± 0.37 17.60 ± 0.87 22.98 ± 0.17 32.05 ± 0.94 46.06 ± 1.33 28.60 ± 2.11 20.22 ± 1.22 37.28 ± 0.96 4.43 ± 0.11 15.65 ± 0.32 61.13 ± 0.70 27.27 ± 0.80 59.42 ± 3.65 — 28.66 ± 5.05 — — — 573.98

55.0 82.0 26.9 15.3 17.0 23.6 33.0 45.2 28.5 19.9 35.4 3.8 14.1 59.2 2.1 36.0 — — — — — 582.2

53 78 28 15 17 23 34 46 28 19 36 4 14 61 26 59 2 35 — — — 578

41 59 28 16 18 25 34 46 29 20 36 4 14 60 26 58 2 35 — 20 12 583

Cys = cysteine. an

= 3, Data from ref. 17. values were recalculated on the basis of a total of 583 residues considering the fact that two Trp residues were completely destroyed and cysteine residues were substantially destroyed during acid hydrolysis. cData from ref. 18. dData from ref. 24. eData from ref. 25. fData from ref. 26. gn = 12. hn = 4. bThe

Woo


165

33. In the PTC-derivatives, cystine+cysteine showed the lowest γ value (0.967), but it could be used for quantitative analysis. However, there is a report that the linearity for PTC-cystine was so poor that it could not be used for quantitative analysis (22). 34. Amino acid compositions of soybean meal determined by the BTC, BZTC, and PTC-derivatives compared to the ion–exchange chromatography were shown in Table 3. 35. In the BTC-derivatives compared to ion–exchange chromatography, most of amino acids showed a deviation of less than 5%, except for histidine (13%) and isoleucine (9%). But BZTC and PTC-derivatives, most of amino acids showed a deviation of more than 5%. Reproducibility was superior in the BTC and BZTCderivatives compared to PTC-derivatives. 36. Table 4 shows the amino acid compositions of whole egg determined by the BTC and PTC-derivatives compared to ion–exchange chromatography. We did not determine BZTC-derivatives, so we could not show the data of the BZTC-derivatives. As in the soybean meal, the data of BTC-derivatives were well matched with those of ion–exchange chromatography. Reproducibility was also superior in the BTC-derivatives as in the soybean meal. 37. Table 5 shows amino acid compositions of BSA recalculated on the basis of a total of 583 residues considering the fact that 2 Trp residues were completely destroyed, and cysteine residues were substantially destroyed during the hydrolysis. 38. It has been shown by sequence analysis that BSA has 17 disulfide bonds (cystine) and one free cysteine (23). In this chapter, the number of BZTC-cystine and BZTC-cysteine residues was 12.5 and 12.4, respectively. It is known that a substantial amount of cysteine and cystine is destroyed during the hydrolysis of protein with 6 M HCl when the hydrolysis was carried out without conversion of these amino acids. Cysteine and cystine determination seems to give unavoidable error, unless they are modified to other compounds, i.e., pyridylethyl-cysteine or cystecic acid, prior to quantification (18,21).

References 1. White, J. A. and Hart, R. T. (1992) Derivatization methods for liquid chromatographic separation of amino acid, in Food Analysis by HPLC (Nollet, L. M. L., ed.), Marcel Dekker, lnc. New York, pp. 53–74. 2. Bidingmeyer, B. A., Cohex, S. A., and Tarvin, T. L. (1984) Rapid analysis of amino acids using precolumn derivatization. J. Chromatogr. 336, 93–104. 3. Heinrikson, R. L. and Meredith, S. C. (1984) Amino acid analysis by reverse-phase high-perfornance liquid chromatography; precolumn derivatization with phenylisothiocyanate. Anal. Biochem. 136, 65–74 4. White, J. A., Hart, R. L., and Fry, L. C. (1986) An evaluation of the Waters Pico-tag system for the amino acid analysis of food materials. J. Auto. Chem. 8, 170–177. 5. Beaver, R. W., Wilson, D. M., Johes, H. M., and Haydon, K. D. (1987) Amino acid analysis in foods and feedstuffs using precolumn phenylisothiocyanate derivatization and liquid chromatography-precolumn study. J. Ass. Off. Anal. Chem. 70, 425–428.

166

Woo

6. Bidingmeyer, B. A., Cohen, S. A., Tarvin, T. L., and Frost, B. (1987) A new, rapid, high-sensitivity analysis of amino acids in food type samples. J. Ass. Off. Anal. Chem. 70, 241–247. 7. Cohen, S. A. and Strydom, D. J. (1988) Amino acid analysis utilizing phenylisothiocyanate derivatives. Anal. Biochem. 174, 1–16. 8. Koop. D. R., Morgan, E. T., Tarr, G. E., and Coon. M. J. (1982) Purification and characterization of a unique isozyme of cytochrome p-450 from live microsomes of ethanol-treated rabbits. J. Biol. Chem. 257, 8472–8480. 9. Roth, M. (1971) Fluorescence reaction for amino acids. Anal. Chem. 43, 880–882. 10. Yaegaki, K., Tonzetich, J., and Ng, A. S. K. (1986) Improved high-performance liquid chromatography method for quantitation of proline and hydroxyproline in biological materials. J. Chromatog. 356, 163–???. 11. Tapuhi, Y., Schmidt, D. E., Lindner, W., and Karger, B. L. (1981) Dansylation of amino acids for high-performance liquid chromatography analysis. Anal. Biochem. 115, 123–129. 12. DeJong, C., Hughes, G. J., Wieringen, E. V., and Wilson, K. J. (1982) Amino acid analysis by high-performance liquid chromatography. An evolution of usefulness of pre-column Dns derivatization. J. Chromatogr. 241, 345–359. 13. Lin, J. K. and Chang, J. Y. (1975) Chromophoric labeling of amino acids with 4dimethylaminoazobenzene-4'-sulfonyl chloride. Anal. Chem. 47, 1634–1638. 14. Knecht, R. and Chang, J. Y. (1986) Liquid chromatograpic determination of amino acids after gas-phase hydrolysis and derivatization with (dimethylamino)azobenzene-sulfonyl chloride. Anal. Chem. 58, 2375–2378. 15. Cohen, S. A. (1990) Analysis of amino acids by liquid chromatography after precolumn derivatization with 4-nitrophenylisothiocyanate. J. Chromatogr. 512, 283–290. 16. Woo, K. L. and Lee, S. H. (1994) Determination of protein amino acids as butylthiocarbamyl derivatives by reversed-phase high-performance liquid chromatography with precolumn derivatization and UV detection. J. Chromatogr. A. 667, 105–111. 17. Woo, K. L. Hwang, Q. C., and Kim, H. S. (1996) Determination of amino acids in the foods by reversed-phase high-preformance liquid chromatography with a new precolumn derivative, butylthiocarbamyl amino acid, compared to the conventional phenylthiocarbamyl derivatives and ion-exchange chromatography. J. Chromatogr. A. 740, 31–40. 18. Woo, K. L. and Ahan, Y. K. (1996) Determination of protein amino acid as benzylthiocarbamyl derivatives compared with phenylthiocarbamyl derivatives by reversed-phase high-performance liquid chromatography, ultraviolet detection and precolumn derivatization. J. Chromatogr. A. 740, 41–50. 19. Woo, K. L. and Lee, D. S. (1995) Capillary gas chromatographic determination of proteins and biological amino acids as N(o)-tert-butyldimethysilyl derivatives. J. Chromatogr. B. 665, 15–25. 20. Greenstein, J. P. and Winitz, M. (1986) Chemistry of Amino Acid, vol. 3. Robert E. Krieger Publish., Malabar, FL, p. 1882.


167

21. Aitken, A., Geisow, M. J., Findlay, J. B. C., Holmes, C., and Yarwoord, A. (1989) Peptide Preparation and Approach. IRC, Oxford, pp. 43–68. 22. Fürst, P., Pollack, L., Graser, T. A., Gldel, H., and Stehle, J. (1990) Appraisal of precolumn derivatization methods for the high-performance liquid chromatographic determination of free amino acids in bilogical materials. J. Chromatogr. 499, 559– 569. 23. Dayhoff, M. O. (1976) Atlas of Protein Sequence and Structure. vol. 5, Suppl. 2, Natl. Biomed. Res. Found., Washington, DC, pp. 267. 24. King, T. P. and Spencer, M. (1970) Structural studies and organic ligand-binding properties of bovine plasma albumin. J. Biol. Chem. 245, 6134–6148. 25. J. R. Brown (1975) Structure of bovine serum albumin. Fed. Proc. 34, 591. 26. Hirayama, K., Akashi, S., Furuya, M., and Fukuhara, K. I. (1990) Rapid confirmation and revision of the primary structure of bovine serum albumin by ESIMS and Frit-FAB LC/MS. Biochem. Biophys. Res. Commun. 173, 639–646.

Amino Acid Measurement

169

12 Amino Acid Measurement in Body Fluids Using PITC Derivatives Roy A. Sherwood 1. Introduction Amino acid chromatography is used to detect both primary disorders of amino acid metabolism (e.g., maple syrup urine disease) and disorders of renal tubular reabsorption (e.g., cystinuria) (see Notes 1–4). In most patients with disorders in the former group, the abnormal amino acids are clearly increased in both plasma and urine, although the abnormality is usually more pronounced in urine. In disorders of renal tubular transport, only the urine amino acids will be abnormal, thus these conditions will be missed if plasma alone is studied. Quantitative amino acids are measured in blood samples from children who are having their amino acids levels manipulated by dietary restriction of natural protein. Derivatization of amino acids is required before analysis by high-performance liquid chromatography (HPLC). A number of alternative pre- or postcolumn derivatization methods have been described (1). In the method described here amino acids in standards, urine, or deproteinized serum/plasma are reacted with phenylisothiocyanate (PITC) in the presence of a coupling solvent. The phenylthiocyanate (PTC) derivatives are then applied to a reversephase (RP) HPLC column and separated using a gradient elution system. The PTC amino acids are detected using an ultraviolet (UV) detector (254 nm) and an electrochemical detector (ECD) in series (2). 2. Materials 2.1. Equipment 1. Gradient HPLC system with ODS Hypersil 5 µm, 4 mm × 25-cm column with precolumn filter. 2. UV detector and amperometric electrochemical detector. From: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

169

170

Sherwood

Table 1 Composition of Coupling Reagent Chemical Methanol (Hipersolv) Triethylamine Distilled water neat PITC

2 samples

10 samples

30 samples

70 µL 10 µL 10 µL 10 µL

350 µL 50 µL 50 µL 50 µL

1050 µL 150 µL 150 µL 150 µL

2.2. Reagents and Solutions 1. Acetic acid 6% (v/v). 2. Buffer A: 10 mM sodium acetate adjusted to pH 6.4 with 6% acetic acid. Should be prepared fresh for each run. Two liters will be enough to run 15 samples. 3. Buffer B: 10 mM acetonitrile, sodium acetate 60% (v/v) adjusted to pH 6.4 with 6% acetic acid. Stable at room temperature for 14 d. One liter will be enough to run 25 samples. Acetonitrile is toxic by ingestion, inhalation, and skin contact. 4. Precipitating reagent containing internal standard: 10% sulphosalicylic acid (SSA), 750 µM norleucine. Stable for 6 mo at 4°C. 5. PITC coupling reagent. Prepared fresh, as per Table 1. PITC is toxic by ingestion, irritating to skin/eyes, and has been reported to have teratogenic effects. Triethylamine is toxic by ingestion and is irritating to the eyes. 6. Amino acid standards (Sigma Chemicals, Poole, UK). Choice dependent on amino acids of interest. Available standards (17 amino acid standard, acids, and neutrals, basics) can be supplemented by specific amino acids as required.

3. Methods

3.1. Preparation of Samples Plasma, random urine (children) or 24 h urine (adults) can be used (see Notes 5 and 6). Heparinized plasma is preferred, there are differences between plasma and serum amino acids. In acutely ill children with a suspected inherited metabolic defect, blood should be taken on admission. If the acute illness occurred some days before investigation, dietary protein may have been withdrawn and an amino acid abnormality may not be apparent. In a nonacute situation, a postabsorptive or fasting sample is preferred (adults and older children fasted overnight, smaller children 6–8 h). This is because plasma amino acids may be significantly increased 1–2 h after a protein meal. In neonates samples should be taken 2–3 h after feeding. 1. Urine. For urine, determine the creatinine in mmol/L (see Note 7). For creatinine concentrations below 5 mmol/L, take 50 µL of urine, make up to 100 µL with distilled water, and multiply result by two. For urine creatinine concentrations


171

above 5 mmol/L, take 20 µL of urine, make up to 100 µL with distilled water, and multiply result by 5. For urine creatinine concentrations below 1 mmol/L, another sample should be requested as the urine is too dilute to give meaningful results. 2. Add 50 µL of SSA precipitating reagent to tube. Mix and leave for 30 min at 4°C. 3. Plasma/serum: Pipet 100 µL of sample into a microfuge tube. Add 50 µL of SSA precipitating reagent to tube. Mix and leave for 30 min at 4°C. 4. Standards: 10 µL added to 90 µL of distilled water. Additional amino acid standards can be added at this point with corresponding reduction in the water, e.g., add 10 µL of a 1 mg/mL standard. Standards as prepared give 250 µmol/L concentrations except cystine (125 µmol/L). Add 50 µL of precipitating reagent containing internal standard to tube. Mix and leave for 30 min at 4°C.

3.2. Derivative Formation 1. Centrifuge tubes for 10 min at 14,000g. 2. Transfer 50 µL of supernatant to an appropriately labeled microfuge tube. Add 200 µL acetonitrile and mix. 3. Add 40 µL PITC/coupling reagent and vortex mix. Leave to react for 20 min. 4. Evaporate to dryness at a temperature less than 45°C. It is essential that the derivatives are dry, but not cooked/desiccated. The derivatives (dry) can be left at this stage at 4°C (no more than 72 h). 5. Dissolve in 200 µL of sodium acetate buffer (Buffer 1). Mix. 6. Centrifuge and remove 200 µL of supernatant to HPLC vial with microinsert for chromatography.

3.3. Chromatography 1. HPLC operating conditions: Column oven temperature: 30°C, sample cooler, 4°C, injection volume 25 µL. Electrochemical detector settings: potential +1.1 V (oxidation), sensitivity 3 µA. UV detector: wavelength 254 nm. If the UV detector and the electrochemical detector are both used in series the UV detector must be placed first. 2. Gradient: The solvent gradient is set as shown in Table 2. 3. Allow the system to equilibrate and run a gradient off-line while preparing samples and standards. The cycle time on the autosampler is 95 min. 4. Inject the first standard in the queue × 2. (The first sample retention times are often too variable for calibration purposes.) 5. Typical retention times and relative (to the internal standard norleucine) retention times are shown in Table 3 (see Notes 8–12). 6. Calculation: To obtain calculation factor: Calculation factor = conc of std (µmol/L) × area/height of Internal Standard area/height of Standard The peak area/height of unknowns is multiplied by this factor to give the concentration in µmol/L.

172

Sherwood

Table 2 HPLC Solvent Gradient Time (min)

Buffer 1 (%) Sodium acetate, pH 6.4

Buffer 2 (%) 60% acetonitrile/ sodium acetate, pH 6.4

Flow rate mL/min

0 20 65 67.5 70 75 80 85 95

100 87 45 0 0 100 100 100 100

0 13 55 100 100 0 0 0 0

1.0 1.0 1.0 2.0 2.0 2.0 1.0 1.0 1.0

7. Interfering substances: Vigabatrin will form a PTC derivative. The cephalosporin group of antibiotics potentially interfere with the UV detection method and might cochromatograph with glycine.

4. Notes 1. In approx 99% of cases, the amino acid pattern will be within normal limits. Variations occur for physiological reasons, particularly in premature neonates, but in comparison to pathological processes the variation is small. Most inherited disorders of amino acid metabolism produce gross changes in the observed pattern. Children diagnosed with inherited diseases involving amino acids are often treated with artificial feeds excluding certain amino acids, and they can get very low levels of essential amino acids, necessitating frequent monitoring. 2. Pathological cases: Most are fairly obvious and the specific amino acid will increase dramatically, e.g., in phenylketonuria, cystinuria, maple syrup urine disease, cystathioninuria, and hypophosphatasia the increase on HPLC will be off scale by 2–10 fold. Homocystinuria is an exception to this in that homocystine is normally undetectable by HPLC. A positive cystine screen with a normal cystine on thin layer chromatography should alert you to the possibility of homocystinuria. Using the HPLC method, it chromatographs immediately after the internal standard and any visible peak must be considered abnormal, but this may not be easy to pick up. Methionine concentrations are often high, secondary to liver disease, such as tyrosinemia and galactosemia. 3. Symptoms suggestive of specific disorders: a. Cataracts. Common in homocystinuria. b. Deafness. Occurs in arginosuccinic aciduria. Gives three peaks on HPLC (two anhydrides and the acid itself).


173

Table 3 Mean Retention Times and Relative Retention Times for the Amino Acids Most Likely to be Found in Body Fluids Amino acid Phosphoserine Aspartic acid Glutamic acid γ-Aminoadipic acid Hydroxyproline Phosphoethanolamine Serine Glycine Asparagine Sarcosine β-Alanine Taurine γ-Aminobutyric acid Citrulline Threonine Alanine β-Aminoisobutyric acid Proline Histidine Carnosine Arginine Methyl histidine α-Aminobutyric acid Anserine Tyrosine Valine Ethanolamine Methionine Cystathionine Cystine Isoleucine Leucine Norleucine Hydroxylysine Phenylalanine Ornithine Tryptophan Lysine

Retention time (min)

Relative retention time

3.82 5.44 6.83 10.35 12.25 12.45 14.52 15.53 15.55 16.91 17.40 19.00 20.47 20.65 21.11 21.75 22.27 23.49 23.91 25.71 28.22 28.39 28.60 29.06 35.02 35.76 37.66 38.10 38.20 40.88 42.79 43.47 45.00 47.78 48.20 49.67 50.16 52.79

0.084 0.12 0.15 0.229 0.271 0.277 0.321 0.345 0.347 0.375 0.387 0.421 0.454 0.458 0.468 0.482 0.494 0.521 0.532 0.569 0.626 0.630 0.634 0.644 0.778 0.795 0.835 0.845 0.848 0.907 0.950 0.965 — 1.060 1.069 1.102 1.112 1.170

174

4.

5.

6.

7.

Sherwood c. Abnormal X-ray. Hypophosphatasia (serum alkaline phosphatase activity will be < 25 IU/L). d. Renal Stones. Cystinuria (also possibly xanthine/purine stones which are radiolucent in disorders such as Lesch-Nyhan syndrome). e. Organic acidurias. Most organic acidurias produce a metabolic acidosis, a urine pH < 6.0 and coma. Some include hypoglycemia and abnormal blood lactate/pyruvate ratio. There are often abnormalities in the amino acid pattern, but these are not always helpful. Physiological/spurious increases in amino acids: a. Tyrosine. Transient neonatal tyrosinaemia is common in the first 6–8 wk of life for premature infants. If it persists after this time, then either tyrosinaemia or liver disease should be suspected (the latter is about 10 times more likely). The cutoff for tyrosine levels giving rise to recall in neonatal screening programs is typically 500 µmol/L. However, recent experience has shown that tyrosine concentrations in tyrosinemia type I can be around 200 µmol/L with a neonatal presentation. b. Threonine. Increased in neonates when compared to adults. It is also increased in liver disease. c. Taurine. Increased taurine can occur in catabolic states, but there is no specific disorder involving taurine. d. Glycine/aspartic acid/glutamic acid. Hyperglycinemia can occur in organic acidurias, valproate treatment, and in subjects with a low calorie intake. In urine samples with bacterial contamination, glutamine is deaminated to glutamic acid, aspartic acid may be increased, and hippuric acid is broken down to glycine. A pH >7.0 is suspicious in such cases. e. Lysine/cystine. Both are often raised in the first 6 mo of life. A generalized amino aciduria is often seen in very sick neonates with poor peripheral circulation. Particular increases are seen in alanine, glycine and proline related to disturbances of lactate/pyruvate metabolism. Samples should not be hemolyzed. Samples are stored frozen if not assayed immediately. Samples left unseparated will have increases in taurine, phosphoethanolamine, aspartic, and glutamic acid. Free cystine and homocystine in plasma should only be measured in samples that have been separated and frozen within 30 min of being taken. Care should be taken not to aspirate the buffy coat with the plasma, this leads to increases in taurine, aspartic acid, glutamic acid, and phosphoethanolamine. A 24-h collection is required for monitoring treatment of cystinuric patients (a compromise for adults is to fast overnight and collect a 4-h (still fasting) collection (from 0630–1030) (3). An aliquot of urine should be taken on receipt for creatinine estimation, the remainder of the urine should be frozen until analyzed. Excretion rates of amino acids may vary independently of the creatinine over a 24-h period. Diurnal variation has been observed.


175

Fig. 1. Formation of the PTC derivatives of amino acids.

8. Identify peaks on samples by cross reference to standards and to relative retention times obtained on previous runs. Peaks can be identified by cochromatography. Mix equal volumes of a derivatized standard and sample, cochromatography and the amino acid is probably the same as in the sample, splitting of the peak — it is not the same. 9. Several of the peaks are pH dependent and move slightly as the buffer ages, in addition, separations tend to vary with different batches of the column. Typically, proline moves to merge with histidine or alanine. Hydroxyproline, ethanolamine, and phosphoethanolamine also move, but this is seldom a problem. 11. Histidine can be identified by its characteristic tailing peak. 12. Threonine and citrulline tend to cochromatograph, citrulline runs slightly fast of threonine. 13. Glycine and asparagine occasionally run together, glycine is faster than asparagine.

Acknowledgments The sulphosalicylic acid precipitation modification of the original method was a personal communication by Dr. Steve Kryawych (Great Ormond Street Hospital). References 1. Walker, V. and Mills, G. A. (1995) Quantitative method for amino acid analysis in biological fluids. Ann. Clin. Biochem. 32, 28–57. 2. Sherwood, R. A., Titheradge, A. C., and Richards, D. A. (1990) Measurement of plasma and urine amino acids by high-performance liquid chromatography with electrochemical detection using phenylisothiocyanate derivatization. J. Chromatog. 528, 293–303. 3. Parvy, P. R., Bardet, J. I., Rabier, D. M., and Kamoun, P. (1988) Age related reference ranges for free amino acids in first morning urine specimens. Clin. Chem. 34, 2092–2095.

Analysis of Proteins and Phospholipids in Lipid-Rich Materials

177

13 Determination of Proteins, Phosphatidylethanolamine, and Phosphatidylserine in Lipid-Rich Materials by Analysis of Phenylthiocarbamyl Derivatives Margareta Stark and Jan Johansson 1. Introduction In classical, ninhydrin-based amino acid analysis (1,2), the ion–exchange matrix used for separation becomes contaminated upon consecutive analyzes of extremely lipid-rich samples; in our experience, already after approx 20–30 samples. Therefore, in analysis of lipid-rich material we focused on amino acid analysis involving reversed phase (RP) chromatography, because lipids are soluble in the organic solvents commonly used for elution and regeneration of such columns (e.g., acetonitrile and 2-propanol). Precolumn derivatization with phenylisothiocyanate (PITC) of protein hydrolysates from physiological samples followed by RP-HPLC is equivalent in analytical quality to the classical ion–exchange chromatography/ninhydrin method (3), and is sensitive down to at least 10 pmol (4). PITC reacts both with primary and secondary amines, the reproducibility is high and the phenylthiocarbamyl (PTC) amino acid derivatives are stable for months when stored dry at –20°C, or for days in solution at ambient temperature (3,4). This approach can be used conveniently for analysis of lipid-rich material, where at least 300 samples can be analyzed with the same column, i.e., the column lifetime is comparable to those encountered during analysis of lipid-free samples (5). During isolation of proteins from lipid-rich sources, we have found it necessary to monitor both protein and phospholipid profiles, at least at early stages of the purification scheme. The phospholipids are usually detected by phosphorous analysis, but with the PTC/RP-HPLC method, phospholipids containFrom: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

177

178

Stark and Johansson

ing a primary amino group can be determined simultaneously with the proteins. Phosphatidylethanolamine subjected to hydrolysis and subsequent analyzed by the PTC/RP-HPLC method show a peak eluting between PTC-Arg and PTC-Tyr, with a linear range between 1–50 nmol, but lower levels are detectable. With the classical ion–exchange/ninhydrin amino acid analysis, where ethanolamine coelutes with Lys and consequently neither of these compounds can be determined. Phosphatidylserine analyzed in the same way by the PTC/ RP-HPLC method gives one peak that coelutes with PTC-Ser, with a linear range between 10 pmol–50 nmol. When monitoring the protein and phospholipid profile after chromatography, it is feasible to determine when serine derived from phosphatidylserine elutes by looking at the overall PTC-amino acid profile and amounts. Phosphatidylcholine contains a quaternary amino group and cannot be detected with the PTC/RP-HPLC method (5). A major advantage with PTC/RP-HPLC analysis, in addition to allowing repeated analysis of lipid-rich samples, is that phosphatidylethanolamine and phosphatidylserine can be determined simultaneously with the proteins, thereby reducing the need for separate phosphorous analysis during protein purification. A sample from a lung phospholipid fraction, which contains about 2% protein and 98% lipids, was hydrolyzed and analyzed by the PTC/RP-HPLC method (Fig. 1). 2. Materials

2.1. Hydrolysis 1. Glass tubes for hydrolysis and derivatization are 6–7 × 35 mm. Tubes for mixing derivatization solutions are 10 × 75 mm. Submit all glass tubes to pyrolysis (400– 500°C for 3–4 h) before use to remove any contaminating material. Recommended glass quality: soda-lime or Duran. 2. Hydrolysis solution: 6 M HCl with 0.5% (w/v) phenol.

2.2. Derivatization 1. Amino acid standard in 0.1 M HCl (e.g., Amino Acid Standard H, Pierce). A stock solution of 100 µM of each amino acid in water can be stored at –20°C for several months. 2. Phosphatidylethanolamine (e.g., dipalmitoylphosphatidylethanolamine, Fluka) dissolved in chloroform, e.g., 10 mM stock solution. Store in capped glass tube at –20°C. 2. Triethylamine (sequanal grade, Pierce). Store at 4°C. NB! Avoid inhalation and skin contact. Wear protective clothing. Work in a ventilated hood and use gloves. 3. Phenylisothiocyanate (PITC) (sequanal grade, Pierce). Store opened ampoule in a screw-capped glass tube under nitrogen in the dark at ambient temperature. Discard when yellow. Caution: Avoid inhalation and contact with skin and eyes. Work in a ventilated hood, use gloves, and wear protective clothing.


179

Fig. 1. RP-HPLC of PTC derivatives from a sample obtained from a phospholipid/ Lipoprotein fraction from porcine lung. The sample was hydrolyzed and analyzed by the PTC/RP-HPLC method. The PTC derivatives are identified with the one-letter code for the corresponding amino acid or EA for ethanolamine. The arrow identifies the peak that appears when dipalmitoyl-phosphatidylethanolamine was hydrolyzed separately.

4. Pre-derivatization solution: 99.5% ethanol/water/triethylamine (2:2:1, v/v/v), 40 µL per sample is required. Mix in a 10 × 75-mm glass tube. Prepare fresh. 5. 50% ethanol: 99.5% ethanol/Milli-Q water (1:1, v/v). Store in a screw-capped glass tube at 4°C. 6. Derivatization solution: 99.5% ethanol/triethylamine/PITC (7:2:1, v/v/v). Mix in a 10 × 75-mm glass tube. Prepare fresh.

2.3. Separation 1. Mobile phase A: 30 mM phosphate buffer, pH 6.30–6.80 (see Note 1). Dissolve 1.2 g NaOH in approx 800 mL MilliQ water, adjust to the pH required with 1 M H3PO4. Adjust to 1000 mL and filter through 0.2 µm. Degas before use. Use fresh or store at 4°C overnight. If stored longer (a few days), filter again through 0.2 µm before use. 2. Mobile phase B: 60% acetonitrile in water. Mix 600 mL acetonitrile with 400 mL Milli-Q water. Store at ambient temperature. Stable for several weeks. Degas before use. 3. Column: C18 reversed phase HPLC column, Spherisorb S3 ODS2, 4.6 × 100 mm, packed with 3-µm particles (Waters).

180

Stark and Johansson Table 1 Gradient Program for Separation of the PTC-Derivatives Time (min)

%A

%B

0 1 26 (Note 6) 28 34 37

100 95 68 (66) 0 0 100

0 5 32 (34) 100 100 0

Flow rate 1.0 mL/min, the gradient is linear. Next sample injected after 8 min delay (45 min total per sample).

4. High-performance liquid chromatography (HPLC) instrumentation: Any analytical instrument equipped with two separate pumps, absorbance detector (254 nm), integrator, heating block, and a temperature control unit.

3. Methods 3.1. Hydrolysis 1. Pipet samples/phospholipid standards (see Note 2) present in volatile solution in 6 × 35-mm glass tubes and dry the samples under a stream of nitrogen or vacuum. For the acid hydrolysis add 40 µL 6 M HCl/0.5% phenol to each tube, evacuate and seal the tubes. Incubate at 110°C for 20–24 h, open and dry the samples under vacuum.

3.2. Derivatization 1. Pipet amino acid standard (1000–5000 nmol/aa) in a 6 × 35-mm glass tube and dry under vacuum (see Note 3). 2. Add 40 µL pre-derivatization solution (99.5% ethanol/water/triethylamine; 2:2:1) (6) to each of the hydrolyzed samples/phospholipid standards and amino acid standard, seal the tubes with parafilm, and mix for 10 s. Remove the parafilm and redry under vacuum (see Note 4). 3. For derivatization add 3 µL 50% ethanol to each tube (7), make sure that an aqueous film is obtained at the bottom of the tube by mixing or turning the tube. Then add 7 µL of derivatization solution (99.5% ethanol/triethylamine/PITC, 7:2:1) and seal the tubes with parafilm. Mix vigorously (vortex) for 10 s and allow the reaction to occur for 15–30 min at ambient temperature. Remove the parafilm and dry the samples under vacuum overnight to remove excess reagent (see Note 5). 4. Seal the tubes with parafilm and store at –20°C, if not immediately subjected to chromatography. The samples can be stored for several months in the freezer.


181

3.3. Separation of PTC-derivatives by HPLC 1. Dissolve the derivatized and dried standard and samples in 100 µL mobile phase A. Filter lipid-containing samples through 0.45 µm. 2. Equilibrate the column with mobile phase A, flow rate 1.0 mL/min, and column temperature 36°C. 3. Inject 5–90 µL standard or sample on the column and carry out gradient elution as shown in Table 1 at a flow rate of 1.0 mL/min and a column temperature of 36°C. Allow the column to equilibrate with mobile phase A for 8 min between each run, i.e., one sample injected every 45 min. 4. Quantify the PTC-amino acids and PTC-ethanolamine by comparison of the peak areas with those of the standard mixtures. 5. Store the column between runs in 60% acetonitrile (see Note 7).

4. Notes 1. Start at pH 6.3–6.4. The elution position of His is pH sensitive. Lower pH — earlier retention time of His, higher pH — later retention time. After several runs His usually elutes close to, or coelutes with Pro, then raise the pH (0.05–0.1 pH U). Do not exceed pH 7.0, the matrix of the column is not stable in basic pH. 2. Phosphatidylethanolamine standard, suggested amounts: hydrolyze 10 nmol and analyze 10% (i.e., 1 nmol) or make a dilution series. Analysis of less than 500 pmol gives unreliable results. A phosphatidylserine standard is not necessary, the serine from the amino acid standard can be used. 3. Do not hydrolyze the amino acid standard. 50 pmol per amino acid analyzed is suitable for AUFS 0.005 at 254 nm. 4. This step is indispensable. PITC do not react with the amino groups if any residual acid from the hydrolysis is present. 5. PITC has low volatility. Remove by high vacuum in a dessicator with solid NaOH for at least 12 h. If the reagent peak is to high, redry for longer time (e.g., 24 h). 6. Lower %B — later retention times from Ser. Higher % B — earlier retention times from Ser. 7. After 30–50 runs, wash the column with methanol/2-propanol, 1/1 (v/v) to minimize lipid contamination. Store the column in at least 60% acetonitrile.

Acknowledgments We are grateful to Carina Palmberg for skillful assistance and Professor Hans Jörnvall for support. References 1. Spackman, D. H., Stein, W. H., and Moore, S. (1958) Automatic recording apparatus for use in chromatography of amino acids. Anal. Chem. 30, 1190–1206. 2. Hamilton, P. B. (1963) Ion exchange chromatography of amino acids. A single column, high resolving, fully automatic procedure. Anal. Chem. 35, 2055–2064.

182

Stark and Johansson

3. Sarwar, G. and Botting, H. G. (1993) Evaluation of liquid chromatographic analysis of nutritionally important amino acids in food and physiological samples. J. Chromatog. 615, 1–22. 4. Bergman, T., Carlquist, M., and Jörnvall, H. (1986) Amino acid analysis by high performance liquid chromatography of phenylthiocarbamyl derivatives, in: Advanced Methods in Protein Microsequence Analysis (Wittmann-Liebold, B., Salnikov, J., and Erdmann, V. A., eds.), Springer, Berlin/Heidelberg, pp. 45–55. 5. Stark, M., Wang, Y., Danielsson, O., Jörnvall, H., and Johansson, J. (1998) Determination of proteins, phosphatidylethanolamine, and phosphatidylserine in organic solvent extracts of tissue material by analysis of phenylthiocarbamyl derivatives. Anal. Biochem. 265, 97–102. 6. Bidlingmeyer, B. A., Cohen, S. A., and Tarvin, T. L. (1984) Rapid analysis of amino acids using pre-column derivatization. J. Chromatog. 336, 93–104. 7. Koop, D. R., Morgan, E. T., Tarr, G. E., and Coon, M. J. (1982) Purification and characterization of a unique isozyme of cytochrome P-450 from liver microsomes of ethanol-treated rabbits. J. Biol. Chem. 257, 8472–8480.

Analysis of O-Phosphoamino Acids in Biological Samples

183

14 Analysis of O-Phosphoamino Acids in Biological Samples by Gas Chromatography with Flame Photometric Detection Hiroyuki Kataoka, Norihisa Sakiyama, Yukizo Ueno, Kiyohiko Nakai, and Masami Makita 1. Introduction The separation and determination of O-phosphoamino acids has been carried out by thin-layer chromatography (1–8), thin-layer electrophoresis (1,6,7,9–14), gel electrophoresis (15), amino acid analyzer (16–18), high-performance liquid chromatography (HPLC) (19–33), capillary zone electrophoresis (CZE) (34,35) and immunoassay (36–38). However, most of these methods require 32P-labeling for detecting and quantifying O-phosphoamino acids, and therefore cannot be used for analysing nonradiolabeled O-phosphoamino acids in biological samples. Furthermore, some of them show poor resolution or low sensitivity. HPLC methods based on the precolumn ultraviolet (UV) derivatization with phenyl isothiocyanate (22–24), dabsyl chloride (25), and N-α-(2,4-dinitro-5-fluorophenyl)-l-alaninamide (26), the postcolumn fluorescence derivatization with o-phthalaldehyde (27–31), the precolumn fluorescence derivatization with 9-fluorenylmethyl chloroformate (32,33), and CZE methods based on the UV derivatization with phenyl isothiocyanate or dabsyl chloride (34,35) were highly sensitive, but some of these methods lack specificity. Furthermore, many of these methods require lengthy separation time, and clean-up of the sample to remove the excess reagent and coexisting substances. On the other hand, the immunoassays by antibodies raised against each O-phosphoamino acid have been employed for measuring endogeneously phosphorylated amino acids, but these methods involve some problems in the specificity and crossreactivity of the antibodies. Current methods for the deterFrom: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

183

184

Kataoka et al.

Fig. 1. Derivatization process of O-phosphoamino acids.

mination of O-phosphoamino acids have also been described in detail in ref. 39, and the review in ref. 40. Recently, we have developed a selective and sensitive method for the determination of O-phosphoamino acids, such as O-phosphoserine (P-Ser), Ophosphothreonine (P-Thr), and O-phosphotyrosine (P-Tyr) by GC with flame photometric detection (FPD) using a capillary column, in which these compounds were analyzed as their N-isobutoxycarbonyl (isoBOC) methyl ester derivatives (41–49). By using this method, we demonstrated that the contents of free O-phosphoamino acids in tissues (42,43) and urine (48), and proteinbound O-phosphoamino acids in phosphorylated proteins (44,45), tissues (46,49), and urine (47–49) could be rapidly and simply analyzed. In this chapter, selective and sensitive methods for the determination of O-phosphoamino acids in biological samples by FPD-GC are described on the basis of the aforementioned results, and optimum conditions and typical problems encountered in the development and application of the methods are discussed. The derivatization process is shown in Fig. 1. N-isoBOC methyl esters of Ophosphoamino acids can be easily, rapidly, and quantitatively prepared by the reaction with isobutyl chloroformate (isoBCF) by shaking in aqueous alkaline media, followed by esterification with diazomethane. The O-phosphoamino acids in peptides or proteins were released from the sample by acid and base hydrolyzes prior to derivatization. The main advantage of this method is that these amino acids can be easily converted in an aqueous medium without any further clean-up procedure, and the derivatives are stable to moisture. Furthermore, the derivatives can be quantitatively and reproducibly resolved as single peaks using a capillary column, and provide an excellent FPD response.


185

2. Materials

2.1. Equipment 1. Shimadzu 14A gas chromatograph equipped with a flame ionization detector and a flame photometric detector (P-filter). 2. For the analysis of free O-phosphoamino acids, fused silica capillary columns (15 m × 0.53-mm id, 1.0-µm film thickness) of crosslinked DB-1701 (14% cyanopropylphenyl-86% methylpolysiloxane, J & W, Folsom, CA) or DB-210 (50% trifluoropropyl-50% methylpolysiloxane, J & W). 3. For the analysis of protein-bound P-Ser/P-Thr and P-Tyr, DB-1707 (15 m × 0.53mm id, 1.0-µm film thickness) and DB-5 (5% phenyl-95% methylpolysiloxane, J & W, Folsom, CA, USA: 15 m × 0.53-mm id, 1.0-µm film thickness). 4. Model LK-21 ultradisperser (Yamato Kagaku, Tokyo, Japan). 5. Model RD-41 Centrifugal evaporator (Yamato Kagaku). 6. Pico-Tag workstation (Waters Associates, Milford, MA). 7. Pasteur capillary pipet (Iwaki glass No. IK-PAS-5P).

2.2. Reagents 1. Standard O-phosphoamino acids, aminophosphonic acids and nonphosphorylated amino acids: O-phospho-L-serine (P-Ser), O-phospho-D,L-threonine (P-Thr), Ophospho-L-tyrosine (P-Tyr), O-phosphoethanolamine (PEA), and 2aminoethylphosphonic acid (AEP) were purchased from Sigma (St. Louis, MO). All of the nonphosphorylated amino acids used were purchased from Ajinomoto (Tokyo, Japan). 2. Internal standards (IS): 2-amino-4-phosphonobutyric acid (APB), 2-amino-7phosphonoheptanoic acid (APH) and 2-amino-8-phosphonooctanoic acid (APO) as internal standards (IS) were purchased from Sigma. 3. Protein samples: purified commercial proteins, bovine serum albumin (Fraction V, 96–99%), hen egg ovalbumin (Type VII), bovine milk α-casein, bovine milk β-casein, bovine milk κ-casein, egg yolk phosvitin, human hemoglobin, bovine heart cytochrome c, salmon sperm protamine, calf thymus histone (Type II), bovine brain myelin basic protein, yast alcohol dehydrogenase, bovine liver catalase, egg white lysozyme, and bovine pancreas chymotrypsin were purchased from Sigma. 4. Derivatizing reagents: isobutyl chloroformate (isoBCF) obtained from Tokyo Kasei Kogyo was used without further purification and stored at 4°C when not in use. N-methyl-N-nitroso-p-toluenesulphonamide and diethyleneglycol monomethyl ether for the generation of diazomethane (50) were obtained from Nacalai Tesuque. 5. Other materials: peroxide-free diethyl ether was purchased from Dojindo Laboratories (Kumamoto, Japan). Distilled water was used after fresh purification with a Model Milli-Q Jr. water purifier (Millipore, Bedford, MA). All other chemicals were analytical grade.

186

Kataoka et al.

Fig. 2. Shematic flow diagram of the derivatization procedure.

2.3. Solutions 1. Standard O-phosphoamino acid solutions: each compound is dissolved in distilled water to make up a stock solution at a concentration of 0.1 mg/mL and then stored at 4°C. The working standard solutions are made up freshly, as required by dilution of the stock solution with distilled water. These solutions are stable at 4°C for at least 2 wk. 2. Internal standard solutions: APB, APH, and APO are dissolved in distilled water to make up stock solutions at a concentration of 0.1 mg/mL. These solutions are stable at 4°C for at least 2 wk. 3. Trichloroacetic acid (TCA) solution: TCA is dissolved in distilled water at concentrations of 5% and 20%.

3. Method 3.1. Derivatization Derivatization procedure of O-phosphoamino acids is shown in Fig. 2 (see Note 1). 1. For the analysis of free O-phosphoamino acids, pipet an aliquot of a sample containing 10–2000 ng of each amino acid and 0.1 mL of 2.5 µg/mL APB (IS) (if necessary) into a 10-mL Pyrex glass reaction tube with a PTFE-lined screw cap. 2. For the analysis of O-phosphoamino acids in protein hydrolysates, pipet an aliquot of a sample containing 10–500 ng of each amino acid and 0.05–0.1 mL of 1


187

Fig. 3. Apparatus for microscale generation of diazomethane.

3. 4. 5. 6.

7.

8.

µg/mL APB (P-Ser/P-Thr analysis) and APH or APO (for P-Tyr analysis) (IS) (if necessary) into a reaction tube. Adjust these mixtures to pH 10.0–12.0 with 2 M NaOH and make the total volume up to 1 mL with distilled water. Immediately add 0.1 mL of isoBCF to the mixtures and shake at 300 cycles/min (up and down) for 10 min at room temperature (see Note 2). Readjust the reaction mixture to pH 1.0–2.0 with 2 M HCl. Extract twice with 3 mL of diethyl ether to remove the excess of reagent. Shake vigorously for 5–10 s by hand, then centrifugation at 2000g for 30 s (see Note 3). The ethereal extract can then be discarded. Saturate the aqueous layer with NaCl and extract twice with 3 mL of diethyl ether containing 10% of isopropanol with vigorous shaking for 5–10 s by hand. After centrifugation at 2000g for 30 s, the organic layers should be transferred into another tube by means of a Pasteur capillary pipet (see Note 4). The pooled ethereal extracts are methylated by bubbling diazomethane, generated according to the microscale procedure (50) (see Note 5). As shown in Fig. 3, a stream of nitrogen is saturated with diethyl ether in the first side-arm test tube and then passed through a diazomethane generating solution containing N-methyl-N-nitroso-p-toluenesulphonamide, diethylene-glycol monomethyl ether, and KOH. The generated diazomethane is carried into the sample tube via the nitrogen stream until a yellow tinge becomes visible. Leave to stand for more than 5

188

Kataoka et al. min at room temperature. The solvents are then removed by evaporation to dryness at 80°C under a stream of dry air. The residue on the walls of the tube are dissolved in 0.05–0.1 mL of ethyl acetate and then 1 µL of this solution is injected into the gas chromatograph (see Note 6).

3.2. Gas Chromatography 1. DB-1701 (14% cyanopropylphenyl-86% methylpolysiloxane, J & W, GC analysis was carried out with a Shimadzu 14A gas chromatograph equipped with a flame ionization detector and a flame photometric detector (P-filter). 2. For the analysis of free O-phosphoamino acids, fused-silica capillary columns (15 m × 0.53-mm id, 1.0-µm film thickness) of crosslinked DB-1701 or DB-210 should be used (see Note 7). 3. For a DB-1701 column, the operating conditions are as follows: column temperature program at 5°C/min from 180°C to 280°C; injection and detector temperature, 290°C. 4. For a DB-210 column, the operating conditions are as follows: column temperature program at 3°C/min from 200°C to 260°C; injection and detector temperature, 270°C. 5. For the analyzes of protein-bound P-Ser/P-Thr and P-Tyr, DB-1701 (15 m × 0.53mm id, 1.0-µm film thickness) and DB-5 (15 m × 0.53-mm id, 1.0-µm film thickness) should be used (see Notes 8 and 9). 6. For the O-phosphoamino acid analysis of phosphorylated protein samples, a DB1701 column is used with the following operating conditions: column temperatures, 210°C (for P-Ser/P-Thr) and 270°C (for P-Tyr); injection and detector temperatures, 260°C (P-Ser/P-Thr) and 290°C (for P-Tyr) (see Note 10). 7. For the P-Ser/P-Thr analyzes of protein-bound urine and tissue samples, a DB1701 column is used with the following operating conditions: column temperature, isothermal at 220°C; injection and detector temperature, 260°C (see Note 11). 8. For the P-Tyr analyzes of protein-bound urine and tissue samples, a DB-5 column should be used with the following operating conditions: column temperature program at 3°C/min from 230°C to 280°C; injection and detector temperature, 290°C (see Note 11). 9. The nitrogen flow rate in each of these GC analyzes is 10 mL/min. 10. Measure the peak heights of each O-phosphoamino acid and the IS. Calculate the peak height ratios against the IS to construct calibration curves (see Notes 12–14).

3.3. Tissue Sample Preparation 3.3.1. Preparation of Tissue Samples for the Free O-Phosphoamino Acid Analysis Preparation of tissue samples for the free O-phosphoamino acid analysis (43): the following species were used in this experiment: cuttlefish (Ommastrephes sloani pacificus), mackerel (Scomber Japonicus), chicken, mouse, and pig (see Note 15).


189

1. After dissection of each animal, remove the organs and chill on ice. Chop the organs up into pieces and then store frozen at –20°C until the assay. 2. Extract the O-phosphoamino acids from a tissue sample within 48 h by the following procedure, the assay being repeated three times. 3. Homogenize each tissue (0.2–1 g) with 5 vol of 5% of TCA using a Model LK-21 ultradisperser (Yamato Kagaku, Tokyo, Japan). 4. Centrifuge at 2000g for 10 min, then reextract the precipitate with 5 vol of 5% of TCA (see Note 16). 5. The combined supernatants are clarified by passage through a glass filter, if necessary, washed three times with two volumes of diethyl ether to remove TCA, and then warmed at 60°C to remove the ether. 6. After cooling, the solution is adjusted to pH 6.0–7.0 by the addition of a few drops of 1 M NaOH and then run through a Dowex 1 × 8 (4 × 0.7-cm id, AcO– form, 100–200 mesh) column (see Note 17). 7. Wash the column with 15 mL of water and then elute with 0.1 M HCl. The initial 9 mL of the eluate is discarded and the following 5 mL of the eluate (pH 1.0–2.0) is collected as the O-phosphoamino acid fraction. To this fraction, add 0.1 mL of 2.5 µg/mL APB (IS) solution, and the resultant mixture is used for the derivatization.

3.3.2. Preparation of Tissue Samples for the Protein-Bound O-Phosphoamino Acid Analysis Preparation of tissue samples for the protein-bound O-phosphoamino acid analysis (46): seven male ICR mice (6 wk old) were used in the experiments (see Notes 18 and 19). 1. Immediately after dissection, remove each organ and store at –20°C until used. Chop up each pooled tissue. An aliquot (approx 0.5 g) should be homogenized with 5 vol of 5% TCA using an ultradisperser. 2. Centrifuge at 2000g for 10 min and reextract the precipitate with 5 vol of 5% TCA. 3. Wash the resulting pellet twice with 5 mL of diethyl ether to remove the TCA, then warm to dryness at 60°C (see Note 20). 4. Weigh the residue, then ground it to a powder, this being called the protein-bound fraction. 5. Hydrolyze an aliquot (5–10 mg) of the powdered protein-bound fraction with acid and base, as described in Subheading 3.4. 6. Adjust the pH of the acid and base hydrolyzates to pH 10.0–12.0 with 2 M NaOH or 10 M HCl, as described in Subheading 3.4. 7. The samples are then ready for derivatization (see Notes 21 and 22).

3.3.3. Preparation of Urine Samples for the Free O-Phosphoamino Acid Analysis Preparation of urine samples for the free O-phosphoamino acid analysis (48). 1. Collect 24-h urine samples from healthy volunteers under toluene and keep frozen, if not analyzed immediately.

190

Kataoka et al.

2. To 3 mL of urine samples, add 1 mL of 20% TCA and mix. 3. Centrifugation at 2000g for 10 min. 4. Wash the supernatant three times with 3 mL of diethyl ether to remove the TCA, and warm at 60°C to remove the ether. 5. After cooling, the solution should be adjusted to pH 6.0–7.0 by the addition of a few drops of 1 M NaOH and then run through a Dowex 1 × 8 (4.X 0.7-cm id, AcO- form, 200–400 mesh) column. Wash the column with 15 mL of water and then elute with 0.1 M HCl. Discard the initial 7.5 mL of the eluate. Collect the following 5 mL of the eluate (pH 1.0–2.0) as the O-phosphoamino acid fraction. 6. To this fraction, add 0.05 mL of 1 µg/mL APB and APO (IS) solutions. 7. The resultant mixtures are used for the derivatization.

3.3.4. Preparation of Urine Samples for the Total O-Phosphoamino Acid Analysis Preparation of urine samples for the total (free plus protein-bound) Ophosphoamino acid analysis (48) (see Notes 23 and 24): 1. Collect, under toluene, 24-h urine samples from healthy volunteers and keep frozen, if not analyzed immediately. 2. Perform acid and base hydrolyzes of urine samples, as described in Subheading 3.4. 3. For the P-Ser/P-Thr analysis, add 0.1 mL of 1 µg/mL APB (IS) to 0.1 mL of the urine sample and evaporated to dryness at 60°C in a centrifugal evaporator. Hydrolyzed the residue with 0.2 mL of 6 M HCl containing 1% phenol in the vapor phase for 2 h at 110°C under vacuum in a Pico-Tag workstation. 4. For the P-Tyr analysis, add 0.05 mL of 1 µg/mL APO (IS) and 1.22 mL of 10.5 M KOH to 3 mL of the urine sample and hydrolyze for 1 h at 130°C in a Pico-Tag workstation. 5. Adjust the resulting acid and base hydrolysates to pH 10.0–12.0 with 2 M NaOH or 10 M HCl, ready for derivatization.

3.4. Protein Sample Preparation 3.4.1 Acid Hydrolysis For P-Ser and P-Thr analysis the proteins are hydrolyzed in an acid (see Notes 25 and 26). 1. Place an aliquot of the protein sample (0.004–2 mg) in a 5 × 50-mm glass test tube, add 50 µL of 20 µM APB (IS). Dry the mixture in a centrifugal evaporator. 2. Hydrolyze the residue with 0.2 mL of 6 M HCl containing 1% phenol in the vapor phase for 2 h at 110°C under vacuum in a Pico-Tag workstation (51). 3. Extract the resulting hydrolysate twice with 0.5 mL of distilled water, the extracts being transferred to another reaction tube (10-mL Pyrex glass tube with a PTFElined screw cap), and then used for derivatization.


191

3.4.2. Base Hydrolysis For P-Tyr analysis, the proteins are hydrolyzed in a base (see Note 27). 1. Place an aliquot of the protein sample (0.2–4 mg) in a 9 × 75-mm polypropylene tube, add 50 µL of 4 µM APH (IS). 2. Add 0.2 mL of 10 M KOH, and make the total volume up to 0.4 mL with distilled water, if necessary. 3. Hydrolyze the mixture by heating for 1 h at 130°C under vacuum in a Pico-Tag workstation. 4. Add 0.2 mL of 8 M HCl to the resulting hydrolysate and transfer the mixture to another reaction tube. This is now ready for derivatization.

4. Notes 1. IsoBCF and diazomethane are used as a derivatizing reagents for the amino group and the carboxyl and phosphoryl groups, respectively (Fig. 1). 2. The N-isobutoxycarbonylation of O-phosphoamino acids with isoBCF proceeds rapidly and quantitatively in aqueous alkaline media (41). This reaction is completed with >50 µL of isoBCF within 5 min by shaking at room temperature. 3. Although diethyl ether has been used as an extraction solvent for N-isoBOC aminocarboxylic acids (52,53), N-isoBOC phosphoamino acids cannot be extracted with this solvent. Therefore, the reaction mixture is washed with diethyl ether under acid condition in order to remove not only the excess reagent, but also the amines, phenols, and aminocarboxylic acids, which often coexist with O-phosphoamino acids in biological samples and derivatize to corresponding Nand O-isoBOC derivatives under same conditions as described in Note 2. 4. In order to extract the remaining N-isoBOC O-phosphoamino acids in aqueous layer, the addition of another solvent to diethyl ether was tested. Among the various solvents tested, isopropanol proved to be the most satisfactory solvent for this purpose, and its optimum concentration was found to be in the range 5–15%. Thus, the N-isoBOC phosphoamino acids in aqueous layer could be quantitatively and selectively extracted into diethyl ether containing 10% isopropanol in acid condition with NaCl saturation. In this procedure, the organic layers should be collected, taking care to avoid aqueous droplets. It was not necessary to complete draw the organic layer in each extraction. 5. The methylation of the ethereal extracts is successfully carried out by bubbling diazomethane. This reaction should be performed in a well-ventilated hood because diazomethane is explosive and toxic. 6. The derivative preparation can be performed within 30 min, and several samples can be treated simultaneously. The N-isoBOC methyl ester derivatives of Ophosphoamino acids are very stable to moisture; therefore, no precaution to exclude moisture is necessary in their handling and storage. No decomposition was observed during GC analysis and even after standing in ethyl acetate for 3 wk at room temperature.

192

Kataoka et al.

Fig. 4. Chromatograms obtained from standard and animal tissue samples for the free O-phosphoamino acid analysis by FPD-GC. (A) Standard (containing 0.25 µg of each amino acid); (B) mackerel spleen (340 mg); (C) mouse brain (744 mg); (D) pig stomach (596 mg). GC conditions: column, DB-1701 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, programmed at 5°C/min from 180°C to 280°C; injection and detector temperatures, 290°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = 2-aminoethylphosphonic acid (AEP); 2 = O-phosphoethanolamine (PEA); 3 = O-phosphothreonine (P-Thr); 4 = O-phosphoserine (P-Ser); 5 = 2-amino-4phosphonobutyric acid (APB) (IS); 6 = O-phosphotyrosine (P-Tyr).

7. The GC analysis was performed with FPD (526-nm interference filter), which is highly selective for phosphorus containing compounds, by using a megabore capillary column and direct (splitless) sample injection system. In this system, the sufficiently inactivated glass insert should be used to avoid tailing peaks of derivatives caused by the adsorption on injection port. 8. In preliminary test for several megabore capillary columns, DB-17, DB-1701, DB-210, and DB-5 gave good separation for O-phosphoamino acids.


193

Fig. 5. Chromatograms obtained from standard and urine samples for the free Ophosphoamino acid analysis by FPD-GC. (A) Standard (containing 50 ng of each amino acid); (B) and (C) urine (3 mL) samples. GC conditions: column, DB-210 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, programmed at 3°C/ min from 200°C to 260°C; injection and detector temperatures, 270°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = 2-aminoethylphosphonic acid (AEP); 2 = Ophosphoethanolamine (PEA); 3 = O-phosphothreonine (P-Thr); 4 = O-phosphoserine (P-Ser); 5 = 2-amino-4-phosphonobutyric acid (APB) (IS); 6 = 2-amino-8phosphonooctanoic acid (APO) (IS); 7 = O-phosphotyrosine (P-Tyr) (from ref. 48 with permission).

9. For the free O-phosphoamino acid analysis, each amino acid gives a single and symmetrical peak from FPD-GC with a DB-1701, and are completely separated within 16 min from other biological phosphorus amino compounds, such as PEA and AEP (Fig. 4A) (43). Furthermore, the free O-phosphoamino acids can be also separated from these compounds using a DB-210 (Fig. 5A) (48). Other phosphate compounds, such as sugar phosphate and nucleic acid phosphate, and nonphosphorus amino acids were not detected by this method. 10. For the P-Ser/P-Thr and P-Tyr analyzes of phosphorylated protein samples, DB1701 was used under different GC conditions, because these samples were separately treated by acid and base hydrolyzes. As shown in Figs. 6A and 7A, P-Ser/ P-Thr and P-Tyr samples were separated within 12 and 8 min, respectively (44).

194

Kataoka et al.

Fig. 6. Chromatograms obtained from a standard and acid hydrolysates of protein samples by FPD-GC. (A) Standard (containing 1 nmol of each compound); (B) αcasein (0.5 mg); (C) phosvitin (0.01 mg); (D) histone (0.8 mg); (E) myelin (0.25 mg). GC conditions: column, DB-1701 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, 210°C; injection and detector temperatures, 260°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = O-phosphothreonine (P-Thr); 2 = Ophosphoserine (P-Ser); 3 = 2-amino-4-phosphonobutyric acid (APB) (IS) (from ref. 44 with permission).

11. DB-1701 and DB-5 proved to be the most satisfactory columns for the P-Ser/PThr and P-Tyr analyzes of urine hydrolysates, respectively. By using these columns, each O-phosphoamino acid elutes separately as single and symmetrical peaks within 10 min (Figs. 8A and 9A) (48). 12. The calibration curves for the free O-phosphoamino acids by FPD-GC were conducted using APB, which showed a similar behavior to other O-phosphoamino acids during the derivatization and was well separated from these amino acids on a chromatogram as the IS. A linear relationship was obtained at the range 0.01–2 mg with the regression lines for P-Ser, P-Thr and P-Tyr being y = 3.938x – 0.040 (r = 0.9984, n = 18), y = 5.183x – 0.028 (r = 0.9983, n = 18), and y = 1.761x –


195

Fig. 7. Chromatograms obtained from a standard and base hydrolysates of protein samples by FPD-GC. (A) Standard (containing 0.2 nmol of each compound); (B) αcasein (2.0 mg); (C) phosvitin (1.3 mg); (D) histone (2.7 mg); (E) myelin (1.0 mg). GC conditions: column, DB-1701 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, 270°C; injection and detector temperatures, 290°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = 2-amino-7-phosphonoheptanoic acid (APH) (IS); 2 = O-phosphotyrosine (P-Tyr) (from ref. 42 with permission).

Fig. 8. Chromatograms obtained from a standard and acid hydrolysates of urine samples by FPD-GC. ; Standard (containing 100 ng of each compound); (B)–(D) urine (0.1 mL) hydrolysates. GC conditions: column, DB-1701 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, 220°C; injection and detector temperatures, 260°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = O-phosphothreonine (P-Thr); 2 = O-phosphoserine (P-Ser); 3 = 2-amino-4-phosphonobutyric acid (APB) (IS) (from ref. 48 with permission).

196

Kataoka et al.

Fig. 9. Chromatograms obtained from a standard and base hydrolysates of urine samples by FPD-GC. (A) Standard (containing 50 ng of each compound); (B) and (C) urine (3 mL) hydrolysates. GC conditions: column, DB-5 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, programmed at 3°C/min from 230°C to 280°C; injection and detector temperatures, 290°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = 2-amino-8-phosphonooctanoic acid (APO) (IS); 2 = O-phosphotyrosine (PTyr) (from ref. 48 with permission). 0.054 (r = 0.9972, n = 18), respectively, where y is the peak height ratio and x is the amount (µg) of each compound. 13. The derivatives provided excellent FPD responses and the minimum detectable amounts of P-Ser, P-Thr, and P-Tyr at a signal three times as high as the noise under our instrumental conditions were approx 50, 40, and 200 pg as injection amounts, respectively. The FPD-GC system described here was over 200 times more sensitive than the FID-GC system. 14. The calibration curves for the analysis of O-phosphoamino acids in protein hydrolysates were conducted using APB (for P-Ser/P-Thr analysis) and APH or APO (for P-Tyr analysis) as the IS (44,48). In each case, a linear relationship was obtained using different GC columns in the range 10–500 ng of each O-phosphoamino acid and the correlation coefficients were above 0.994 (48). The detection limits for P-Ser, P-Thr, and P-Tyr in these GC conditions were approx 40, 30, and 80 pg as injection amounts, respectively. 15. The method developed for free O-phosphoamino acids was successfully applied to animal tissue (43) and urine (48) samples. In preliminary experiments on mice,


197

Fig. 10. Chromatograms obtained from (A) a standard solution, and (B) nonhydrolyzed and (C) hydrolyzed samples of porcine liver. The converted samples were analyzed by both FID-GC (top) and FPD-GC (bottom). GC conditions: column, DB-1701 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, isothermal at 180°C for 6 min, programmed at 20°C/min to 220°C, isothermal at 220°C for 6 min, programmed at 20°C/min to 270°C, and isothermal at 270°C for 5 min; injection and detector temperatures, 280°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = threonine (Thr); 2 = serine (Ser); 3 = 2-aminoethylphosphonic acid (AEP); 4 = O-phosphoethanolamine (PEA); 5 = O-phosphothreonine (P-Thr); 6 = O-phosphoserine (P-Ser); 7 = tyrosine (Tyr); 8 = O-phosphotyrosine (P-Tyr) (from ref. 42 with permission). no postmortem change of O-phosphoamino acids was observed after 24 h. Ophosphoamino acids in various organs were stable for at least 2 d under freezing at –20°C. 16. TCA is routinely used to precipitate proteins with no loss of phosphoryl groups (54). Free O-phosphoamino acids in tissue and urine samples were quantitatively extracted into the TCA soluble fraction.

198

Kataoka et al.

Table 1 Recoveries of O-Phosphoamino Acids Added to Tissue and Urine Samples Sample Pig tissue Brain

Heart

Liver

Kidney

Urine A

B

Urine hydrolysated A

B

Amount foundb

O-Phosphoamino acid

Addeda

Nonaddition

Addition

Recovery (%)

P-Ser P-Thr P-Tyr P-Ser P-Thr P-Tyr P-Ser P-Thr P-Tyr P-Ser P-Thr P-Tyr

500 100 100 500 100 100 500 100 100 500 100 100

330 ± 20 57 ± 2 NDc 161 ± 11 18 ± 1 ND 377 ± 13 115 ± 2 ND 567 ± 27 46 ± 3 ND

850 ± 14 152 ± 3 103 ± 4 671 ± 50 117 ± 3 97 ± 2 878 ± 64 214 ± 4 102 ± 3 1085 ± 57 146 ± 8 103 ± 5

104.0 95.0 103.0 102.0 99.0 97.0 100.2 99.0 102.0 103.6 100.0 103.0

P-Ser P-Thr P-Tyr P-Ser P-Thr P-Tyr

20 20 20 20 20 20

26.8 ± 0.7 ND ND 12.3 ± 0.5 ND ND

46.4 ± 2.1 19.6 ± 0.6 19.0 ± 1.1 31.8 ± 0.9 18.4 ± 0.7 18.7 ± 1.5

98.0 98.0 95.0 97.5 92.0 93.5

P-Ser P-Thr P-Tyr P-Ser P-Thr P-Tyr

500 500 50 500 500 50

1017 ± 42 133 ± 9 2.4 ± 0.1 1636 ± 60 238 ± 7 6.2 ± 0.4

1504 ± 54 593 ± 35 47.7 ± 0.7 2111 ± 118 694 ± 25 53.5 ± 1.7

97.4 92.0 90.6 95.0 91.2 94.6

aTissue

sample, ng/g; urine sample, ng/mL. sample, ng/g; urine sample, ng/mL; mean ± SD (n=3). cNot detectable. dP-Ser and P-Thr were added to acid hydrolysate and P-Tyr was added to base hydrolysate. bTissue

17. The TCA extract was applied on a Dowex 1 column in order to remove any coexisting substances. O-phosphoamino acids were adsorbed on the column and were eluted in 0.1 M HCl fraction. This column chromatographic procedure could be completed within 50 min, and the recoveries of P-Ser, P-Thr, and P-Tyr were 97– 103%. Figs. 4B–D (43) and 5B,C (48) show the chromatograms obtained from


199

Table 2 O-Phosphoamino Acid Contents in Several Commercial Proteins O-Phosphoamino acid (nmol/mg of protein) a Protein Albumin Ovalbumin α-Casein β-Casein κ-Casein Phosvitin Hemoglobin Cytochrome c Protamine Histone Myelin basic protein Alcohol dehydrogenase Catalase Lysozyme Chymotrypsin

P-Ser b ND d 7.74 ± 0.51 137.30 ± 14.20 104.28 ± 4.87 67.26 ± 0.89 848.14 ± 35.81 0.03 ± 0.002 0.08 ± 0.002 1.46 ± 0.07 0.30 ± 0.01 3.93 ± 0.28 0.07 ± 0.004 0.17 ± 0.01 0.05 ± 0.003 0.13 ± 0.01

P-Thr b

P-Tyr c

ND ND 0.69 ± 0.04 0.16 ± 0.01 0.87 ± 0.01 11.06 ± 0.75 ND ND ND 0.04 ± 0.002 1.00 ± 0.06 ND 0.04 ± 0.001 ND ND

ND ND 0.073 ± 0.004 0.013 ± 0.001 0.071 ± 0.002 0.182 ± 0.010 ND ND ND ND ND ND ND ND ND

± SD (n = 3). after partial acid hydrolysis. cAnalyzed after partial base hydrolysis. dNot detectable. (From ref. 44 with permission.) aMean

bAnalyzed

animal tissue and urine samples, respectively. O-phosphoamino acids could be detected without any interference from coexisting substances by FPD-GC. 18. P-Ser and P-Thr were found in various animal tissues, but only P-Ser was found in urine samples. The occurrence of free P-Ser and P-Thr in these samples was confirmed by GC, GC-mass spectrometry (GC-MS), and thin-layer chromatography of nonhydrolyzed and hydrolyzed samples (42). As shown in Fig. 10, peaks corresponding to P-Ser and P-Thr were observed in the chromatograms obtained by FID-GC (top), and FPD-GC (bottom) analyzes of the porcine liver sample. But these peaks disappeared after acid hydrolysis of sample with 6 M HCl at 110°C for 48 h, and the peaks corresponding to Ser and Thr newly appeared instead (Fig. 10). As shown in Table 1, the recoveries of P-Ser, P-Thr, and P-Tyr added to tissue and urine samples were 92–104%, the relative standard deviations being 1.6–8.0% (n = 3). 19. The method developed for protein-bound O-phosphoamino acids was successfully applied to purified proteins (44), phosphorylated proteins by protein kinases (45), animal tissue (46) and urine (48) samples. For the presence of O-phosphoamino acid residues in protein samples, several commercial proteins were analyzed by FPD-GC. As shown in Figs. 6B–D and 7B–D, O-phosphoamino

200

20.

21.

22.

23. 24.

25.

26.

27.

Kataoka et al. acids in protein hydrolysates were selectively detected without any interference from other amino acids (44). This method is reproducible and directly applicable for the analysis of the extent of both in vivo and in vitro phosphorylation without radiolabeling. The proteins containing O-phosphoamino acids were recovered by better than 99% in the pellet with the TCA treatment without loss of phosphoryl groups (54). Phospholipids, such as phosphatidylcholine and phosphatidylserine, were removed from the protein-bound fraction by washing the pellet with ether. The results of O-phosphoamino acid analyzes of the protein samples after partial acid and base hydrolyzes are shown in Table 2. P-Ser was detected in all the samples investigated in this study, except for BSA, and P-Thr and P-Tyr were also detected in casein and phosvitin. This method was successfully applied to O-phosphoamino acid analysis of the proteins phosphorylated by protein kinases (45). Although the recovery of O-phosphoamino acids from the protein-bound fraction under the best conditions for acid hydrolysis was low (approx 35–40%) (43), the method could reproducibly measure endogeneously phosphorylated amino acids in the tissue and urine samples without radiolabeling. As shown in Table 1, the recoveries of O-phosphoamino acids added to urine hydrolysates were 91–97%, the relative standard deviations being 1.5–6.8% (n = 3). Figs. 8B–D and 9B,C show the chromatograms obtained from acid and base hydrolysates of urine samples by FPD-GC. The P-Ser, P-Thr, and P-Tyr peaks obtained from urine hydrolysates were confirmed by GC-MS analysis (47). By using this method, we demonstrated that P-Tyr levels in mouse urine and liver increase during liver regeneration after partial hepatectomy (49). The O-phosphate linkages of Ser and Thr residues are defined by their stability to acids and lability to bases. Thus, acid hydrolysis has been routinely used to study the identity and amounts of P-Ser and P-Thr in peptides and proteins. For the acid hydrolysis of phosvitin (containing P-Ser, P-Thr, and P-Tyr residues) with the HCl vapor phase using a Pico-Tag workstation, the maximal recoveries of P-Ser and P-Thr were obtained at 110°C for 2–3 h (Fig. 11), these conditions being in agreement with those reported previously (20). Under these conditions, P-Ser was detected as over 70 times more than P-Thr in phosvitin, whereas P-Tyr was not detected at all. Under the best conditions for acid hydrolysis, the recovery of P-Ser from phosvitin was calculated to be 35.7 ± 1.5% (n = 3) by measuring the total phosphate content. The recovery of P-Ser from α-casein was 39.5 ± 4.1% (n = 3) under the same conditions. These low recoveries seem to have been caused by both incomplete peptide hydrolysis and destruction of the liberated P-Ser to Ser. The typical recovery for proteins has been reported to be in the range 20–30% (19). Our results presented here are generally somewhat higher, the recoveries may show some dependence on the neighboring amino acid residues (55). The phosphate linkage of the Tyr residue is far more resistant to base than to acid hydrolysis. P-Tyr tolerated the base hydrolysis fairly well in 5 M KOH at 155°C


201

Fig. 11. Release of O-phosphoamino acids from phosvitin by acid hydrolysis. for 30–35 min (17,54). In our experiment, the maximal recovery of P-Tyr from phosvitin was obtained in 5 M KOH at 130°C for 0.5–4 h (Fig. 12).

References 1. Manai, M. and Cozzone, A. J. (1982) Two-dimentional separation of phosphoamino acids from nucleoside monophosphates. Anal. Biochem. 124, 12–18. 2. Chang, W.-C., Lee, M. L., Chou, C. K., and Lee, S. C. (1983) Polyamide thin-layer chromatography of phosphorylated tyrosine, threonine, and serine. Anal. Biochem. 132, 342–344.

202

Kataoka et al.

Fig. 12. Release of O-phosphoamino acids from phosvitin by base hydrolysis. 3. Fazekas, S., Ovary, I., and Szekessy-Hermann, V. (1989) Presence of phosphotyrosine in alkaline hydrolysate of pig skeletal muscle myosin. Acta. Phys. Hung. 74, 161–168. 4. Neufeld, E., Goren, H. J., and Boland, D. (1989) Thin-layer chromatography can resolve phosphotyrosine, phosphoserine, and phosphothreonine in a protein hydrolyzate. Anal. Biochem. 177, 138–143. 5. Munoz, G. and Marshall, S. H. (1990) An alternative method for a fast separation of phosphotyrosine. Anal. Biochem. 190, 233–237.


203

6. Duclos, B., Marcandier, S., and Cozzone, A. J. (1991) Chemical properties and separation of phosphoamino acids by thin-layer chromatography and/or electrophoresis. Methods Enzymol. 201, 10–21. 7. de Witte, P. A., Cuveele, J. F., Merlevede, W. J., and Vandenheede, J. R. (1995) Analysis of phosphorylhydroxyamino acids present in hydrolyzed cell extracts using dabsyl derivatization. Anal. Biochem. 226, 1–9. 8. de Witte, P. A., Cuveele, J. F., Merlevede, W. J., and Agostinis, P. M. (1996) Analysis of the phosphoamino acid content of phosphoproteins. J. Pharm. Biomed. Anal. 14, 1063–1067. 9. Kamps, M. P. and Sefton, B. M. (1989) Acid and base hydrolysis of phosphoproteins bound to immobilon facilitates analysis of phosphoamino acids in gel-fractionated proteins. Anal. Biochem. 176, 22–27. 10. Hildebrandt, E. and Fried, V. A. (1989) Phosphoamino acid analysis of protein immobized on polyvinylidene difluoride membrane. Anal. Biochem. 177, 407–412. 11. Lippmann, C., Lindschan, C., and Erdmann, V. A. (1992) Thin-layer electrophoresis with PhastSystem facilitates analysis of phosphoamino acids from proteins bound to Immobilon. Electrophoresis 13, 666–668. 12. Durocher, Y. and Chevalier, S. (1994) Detection of phosphotyrosine in partial acid hydrolysis in gels. J. Biochem. Biophys. Methods 28, 101-113. 13. Lombardini, J. B. and Props, C. (1995) Effects of cooling temperature on the separation of phosphoamino acids. Anal. Biochem. 227, 399–400. 14. Mahoney, C. W., Nakanishi, N., and Ohashi, M. (1996) Phosphoamino acid analysis by semidry electrophoresis on cellulose thin-layer plates using the pharmacia/ LKB multiphor or Atto flatbed apparatus. Anal. Biochem. 238, 96–98. 15. Yan, J. X., Packer, N. H., Tonella, L., Ou, K., Wilkins, M. R., Sanchez, J.-C., et al. (1997) High sample throughput phosphoamino acid analysis of proteins separated by one-and two-dimentional gel electrophoresis. J. Chromatog. A 764, 201–210. 16. Kinnier, W. J. and Wilson, J. E. (1977) Complete analysis of protein hydrolysates containing phosphoserine and phosphothreonine using the amino acid analyzer. J. Chromatog. 135, 508–510. 17. Martensen, T. M. (1982) Phosphotyrosine in proteins: Stability and quantification. J. Biol. Chem. 257, 9648–9652. 18. Capony, J. P. and Demaille, J. G. (1983) A rapid microdetermination of phosphoserine, phosphothreonine, and phosphotyrosine in proteins by automatic cation exchange on a conventional amino acid analyzer. Anal. Biochem. 128, 206–212. 19. Morrice, N. and Aitken, A. (1985) A simple amd rapid method of quantitative analysis of phosphoamino acids by high-performance liquid chromatography. Anal. Biochem. 148, 207–212. 20. Robert, J. C., Soumarmon, A., and Lewin, M. J. M. (1985) Determination of Ophosphothreonine, O-phosphoserine, O-phosphotyrosine and phosphate by highperformance liquid chromatography. J. Chromatog. 338, 315-324. 21. McCroskey, M. C., Colca, J. R., and Pearson, J. D. (1988) Determination of [32P]phosphoamino acids in protein hydrolysates by isocratic anion-exchange highperformance liquid chromatography. J. Chromatog. 442, 307–315.

204

Kataoka et al.

22. Meyer, H. E., Swiderek, K., Hoffmann-Posorske, E., Korte, H., and Heilmeyer, L. M. G., Jr. (1987) Quantitative determination of phosphoserine by high-performance liquid chromatography as the phenylthiocarbamyl-S-ethylcycteine: application to picomolar amounts of peptides and proteins. J. Chromatog. 397, 113–121. 23. Murthy, L. R. and Iqbal, K. (1991) Measurement of picomoles of phosphoamino acids by high-performance liquid chromatography. Anal. Biochem. 193, 299–305. 24. Aebersold, R., Watts, J. D., Morrison, H. D., and Bures, E. J. (1991) Determination of the site of tyrosine phosphorylation at the low picomole level by automated solid-phase sequence analysis. Anal. Biochem. 199, 51–60. 25. Malencik, D. A., Zhao, Z., and Anderson, S. R. (1990) Determination of dityrosine, phosphotyrosine, phosphothreonine, and phosphoserine by high-performance liquid chromatography. Anal. Biochem. 184, 353–359. 26. Goodnough, D. B., Lutz, M. P., and Wood, P. L. (1995) Separation and quantification of D-and L-phosphoserine in rat brain using Na-(2, 4-dinitro-5-fluorophenyl)L-alaninamide (Marfey’s reagent) by high-performance liquid chromatography with ultraviolet detection. J. Chromatog. B, 672, 290–294. 27. Steiner, A. W., Helandes, E. R., Fujitaki, J. M., Snith, L. S., and Smith, R. A. (1980) High-performance liquid chromatography of acid-stable and acid-labile phosphoamino acids. J. Chromatog. 202, 263–269. 28. Yang, J. C., Fujitaki, J. M., and Smith, R. A. (1982) Separation of phosphohydroxyamino acids by high-performance liquid chromatography. Anal. Biochem. 122, 360–363. 29. Caelomango, L., Huebner, V. D., and Matthews, H. R. (1985) Rapid separation of phosphoamino acids including the phosphohistidines by isocratic high-performance liquid chromatography of the orthophthalaldehyde derivatives. Anal. Biochem. 149, 344–348. 30. McCourt, D. W., Lykam, J. F., and Schwartz, B. D. (1985) Analysis of sulfate and phosphate esters of amino acids by ion-exchange chromatography on polymeric DEAE. J. Chromatog. 327, 9–15. 31. Etheredge, R. W., III, and Glimcher, M. J. (1986) Resolution and identification of O-phosphoserine, O-phosphothreonine, O-phosphotyrosine, and γ-carboxyglutamic acid as their fluorescent o-phthalaldehyde derivatives by high-performance liquid chromatography. Calcif. Tissue Int. 39, 239–243. 32. Niedbalski, J. S. and Ringer, D. P. (1986) Separation and quantitative analysis of Olinked phosphoamino acids by isocratic high-performance liquid chromatography of the 9-fluorenylmethyl chloroformate derivatives. Anal. Biochem. 158, 138–145. 33. Ringer, D. P. (1991) Separation of phosphotyrosine, phosphoserine, and phosphothreonine by high-performance liquid chromatography. Methods Enzymol. 201, 3–10. 34. Herber, M., Liedtke, C., Korte, H., Hoffmann-Posorske, E., Donella-Deana, A., Pinna, L. A., et al. (1992) Non-radioactive determination of PTH and dabsyl phosphoamino acids by capillary electrophoresis. Chromatographia 33, 347–350. 35. Meyer, H. E., Eisermann, B., Heber, M., Hoffmann-Posorske, E., Korte, H., Weigt, C., et al. (1993) Strategies for nonradioactive methods in the localization of phosphorylated amino acids in proteins. FASEB J. 7, 776–782.


205

36. Wang, J. Y. J. (1988) Antibodies for phosphotyrosine: analytical and preparative tool for tyrosyl-phosphorylated proteins. Anal. Biochem. 172, 1–7. 37. Levine, L., Gjika, H. B., and Vunakis, H. V. (1989) Antibodies and radioimmunoassaya for phosphoserine, phosphothreonine and phosphotyrosine. J. Immunol. Methods 124, 239–249. 38. Heffetz, D., Fridkin, M., and Zick, Y. (1989) Antibodies directed against phosphothreonine residues as potent tools for studying protein phosphorylation. Eur. J. Biochem. 182, 343–348. 39. Hunter, T. and Sefton, B. M. (eds.) (1991) Protein phosphorylation part B, in Methods in Enzynology, vol. 201, Academic, London and New York. 40. Aitken, A. and Learmonth, M. (1997) Analysis of sites of protein phosphorylation. Methods Mol. Biol. 64, 293–306. 41. Kataoka, H., Sakiyama, N., and Makita, M. (1988) Gas chromatographic analysis of aminoalkylphosphonic acids and aminoalkyl phosphates. J. Chromatog. 436, 67–72. 42. Kataoka, H., Sakiyama, N., and Makita, M. (1990) Occurrence of free Ophosphoserine and O-phosphothreonine in porcine liver. Agric. Biol. Chem. 54, 1731–1733. 43. Kataoka, H., Sakiyama, N., and Makita, M. (1991) Distribution and contents of free O-phosphoamino acids in animal tissues. J. Biochem. 109, 577–580. 44. Kataoka, H., Ueno, Y., and Makita, M. (1991) Analysis of O-phosphoamino acids in proteins by gas chromatography with flame photometric detection. Agric. Biol. Chem. 55, 1587–1592. 45. Kataoka, H., Ueno, Y., and Makita, M. (1992) O-Phosphoamino acid analysis of phosphorylated proteins by gas chromatography with flame photometric detection. J. Pharm. Biomed. Anal. 10, 365–369. 46. Kataoka, H., Nakai, K., Ueno, Y., and Makita, M. (1992) Analysis of Ophosphoamino acids in the protein fractions of mouse tissue by gas chromatography. Biosci. Biotech. Biochem. 56, 1300–1301. 47. Kataoka, H., Nakai, K., and Makita, M. (1993) Identification of O-phosphoamino acids in urine hydrolysate by gas chromatography-mass spectrometry. J. Chromatog. 615, 136–141. 48. Kataoka, H., Nakai, K., Katagiri, Y., and Makita, M. (1993) Analysis of free and bound O-phosphoamino acids in urine by gas chromatography with flame photometric detection. Biomed. Chromatogr. 7, 184–188. 49. Kataoka, H., Nakai, K., and Makita, M. (1994) Increase of phosphotyrosine levels in mouse urine and liver during liver regeneration after partial hepatectomy. Biochem. Biophys. Res. Commun. 201, 909–916. 50. Schlenk, H. and Gellerman, J. L. (1960) Esterification of fatty acids with diazomethane on a small scale. Anal. Chem. 32, 1412–1414. 51. Bidlingmeyer, B. A., Cohen, S. A., and Tarvin, T. L. (1984) Rapid analysis of amino acids using pre-column derivatization. J. Chromatog. 336, 93–104. 52. Makita, M., Yamamoto, S., and Kono, M. (1976) Gas-liquid chromato-graphic analysis of protein amino acids as N-isobutoxycarbonylamino acid methyl esters. J. Chromatog. 120, 129–140.

206

Kataoka et al.

53. Yamamoto, S., Kiyama, S., Watanabe, Y., and Makita, M. (1982) Practical gasliquid chromatographic method for the determination of amino acids in human serum. J. Chromatog. 233, 39–50. 54. Martensen, T. M. (1984) Chemical properties, isolation, and analysis of O-phosphates in proteins. Methods Enzymol. 107, 3–23. 55. Bylund, D. B. and Huang, T. S. (1976) Decomposition of phosphoserine and phosphothreonine during acid hydrolysis. Anal. Biochem. 73, 477–485.

Gas Chromatography of Sulfur Amino Acids

207

15 Determination of Sulfur Amino Acids, Glutathione, and Related Aminothiols in Biological Samples by Gas Chromatography with Flame Photometric Detection Hiroyuki Kataoka, Kiyomi Takagi, Hirofumi Tanaka, and Masami Makita 1. Introduction The determination of sulfur amino acids, glutathione (GSH), and related aminothiols has been carried out by isotachophoresis (1), amino acid analyzer (AAA) (2,3), gas chromatography (GC) (4–8), high-performance liquid chromatography (HPLC) (9–32), GC-mass spectrometry (GC-MS) (33,34), liquid chromatography-mass spectrometry (LC-MS) (35), and capillary zone electrophoresis (CZE) (36–38). However, isotachophoresis and AAA methods are nonselective for sulfur amino acids and lack sensitivity. GC methods based on the conversion into trimethylsilyl (4,5), neopentylidine (6), and Ntrifluoroacetyl n-butyl ester (7) derivatives lack sensitivity, gives a tailing peak and requires anhydrous derivatization conditions. Although GC method based on the preparation of N-heptafluorobutyryl isobutyl or ethyl esters (8) is selective and sensitive by flame photometric detection (FPD), this method is not applied to the analysis of biological samples. As the HPLC methods, the ultraviolet (UV) (9–12), and the postcolumn UV derivatization with 4,4'dithiopyridine (13) and 5,5-dithiobis (2-nitrobenzoic acid) (14), the precolumn fluorescence derivatization with 4-(aminosulfonyl)- or ammonium-7-fluoro2,1,3-benzoxadiazole-4-sulfonate (15,16), methyl 4-(6-methoxynaphthalene2-yl)-4-oxo-2-butenoate (17), N-(1-pyrenyl)maleimide (18), monobromobimane (19–23), o-phthaldialdehyde (24,25), 9-fluorenylmethyloxycarbonyl chloride (26), and 2-chloro-1-methylpyridinium (27), and electrochemiFrom: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

207

208

Kataoka et al.

cal detection (28–32) have been reported. Many of these HPLC methods were highly sensitive, but some of these methods lack specificity and require cleanup of the sample to remove the excess reagent and coexisting substances. GCMS methods based on the conversion into tert-butyldimethylsilyl (33) and N(O,S)-propoxycarbonyl propyl ester (34) derivatives, and LC-MS method were highly sensitive and specific, but these methods require expensive equipment. Furthermore, CZE methods with electrochemical detection are capable of achieving higher separation efficiency, use less organic solvents, and require small amounts of samples in comparison with HPLC, but these methods are not applied enough to the analysis of biological samples. Current methods for the determination of sulfur amino acids, GSH, and related aminothiols have also been described in detail in refs. 39–43. Recently, we have developed a selective and sensitive method for the determination of sulfur amino acids by FPD-GC using a DB-17 capillary column (15 m × 0.53-mm id), in which these compounds were analyzed as their N(S)isopropoxycarbonyl (isoPOC) methyl ester derivatives (44). By using this method, we demonstrated that the contents of these amino acids in urine (44) and plasma (45) samples could be rapidly and simply determined. Furthermore, we have developed a selective and sensitive method for the determination of GSH and related aminothiols such as cysteine (Cys), cysteinylglycine (CysGly), and γ-glutamylcysteine (γ-GluCys) by FPD-GC, using a short capillary column (5 m × 0.53-mm id) of crosslinked DB-1 (46). By using this method, we demonstrated that the contents of these aminothiols in blood (46) and tissue (47) samples could be rapidly and simply determined. In the aforementioned methods, other sulfur containing amino acids such as taurine and related aminosulfonic acids could not be analyzed, but these compounds could be analyzed as their N-isoBOC di-n-butylamide derivatives by GC (48–52). In this chapter, selective and sensitive method for the determination of sulfur amino acids, GSH, and related aminothiols, except for aminosulfonic acids by FPD-GC, is described on the basis of the above results, and optimum conditions and typical problems encountered in the development and application of the method are discussed. The derivatization process of sulfur amino acids, GSH, and related aminothiols and structures of main compounds are shown in Fig. 1. N(S)isoPOC methyl esters of these compounds can be easily, rapidly, and quantitatively prepared by the reaction with isopropyl chloroformate (isoPCF) by shaking in aqueous alkaline media, followed by esterification with hydrogen chloride-methanol (HCl-MeOH). The oxidized and protein-bound forms of these compounds were reduced to free thiols by adding sodium borohydride (NaBH4) prior to derivatization. The main advantage of this method is that these amino acids can be easily converted in an aqueous medium without any


209

Fig. 1. Derivatization process and structures of sulfur amino acids, glutathione, and related aminothiols.

further clean-up procedure and the derivatives are stable to moisture. Furthermore, the derivatives can be quantitatively and reproducibly resolved as single peaks using a capillary column, and provide an excellent FPD response. 2. Materials 2.1. Equipment 1. Shimadzu 12A gas chromatograph equipped with a hydrogen flame ionization detector and a flame photometric detector (S-filter). 2. For the analysis of sulfur amino acids, a fused-silica capillary column (15 m × 0.53-mm id, 1.0-µm film thickness) of crosslinked DB-17 (50% phenyl-50% methylpolysiloxane, J & W, Folsom, CA). 3. For the analysis of GSH and related aminothiols, a fused-silica capillary column (5 m × 0.53-mm id, 1.5-µm film thickness) of crosslinked DB-1 (methylpolysiloxane, J & W). 4. Model LK–21 ultradisperser (Yamato Kagaku, Tokyo, Japan). 5. Pasteur capillary pipet (Iwaki glass no. IK-PAS-5P).

2.2. Reagents 1. Amino acids and peptides: L-methionine (Met), ethionine (Eth), L-thioproline (TPro), L-cysteine (Cys), D,L-homocysteine (HCys), L-cystine (Cyt), D,L-ho-

210

Kataoka et al.

mocystine (HCyt), and djenkolic acid (Dje) were purchased from Nacalai Tesque (Kyoto, Japan). L-Cystathionine (CTH), L-methionine sulfone (Met-S), S-methyl-L-cysteine (SM-Cys), S-carboxymethyl L-cysteine (SC-Cys) and cysteinylglycine (CysGly) were purchased from Sigma (St. Louis, MO). Glutathione (GSH) and γ-glutamylcysteine (γ-GluCys) were purchased from Kohjin Co., Ltd. (Tokyo, Japan). Glutathione disulphide (GSSG) was purchased from Wako Pure Chemical Industries (Osaka, Japan). All of the nonsulfur protein amino acids used were purchased from Ajinomoto (Tokyo, Japan). 2. Internal standards: S-2-Aminoethyl-L-cysteine (AE-Cys) and lanthionine (LTH) as internal standard (IS) were purchased from Sigma. 3. Derivatizing reagents: isopropyl chloroformate (isoPCF) was obtained from Wako Pure Chemicals Industries (Osaka, Japan). Hydrogen chloride in methanol (HClMeOH) obtained from Tokyo Kasei Kogyo (Tokyo, Japan) was diluted with methanol at a concentration of 1 M. 4. Other materials: Sodium borohydride (NaBH4) and dithioerythritol (DTE) were obtained from Nacalai Tesque. Peroxide-free diethyl ether was obtained from Dojindo Laboratories (Kumamoto, Japan). Distilled water was used after freshly purification with a Model Milli-Q Jr. water purifier (Millipore, Bedford, MA). All other chemicals were of analytical-reagent grade.

2.3. Solutions 1. Standard sulfur amino acid solutions: three standard stock solutions (each 2 mM), one containing the 13 sulfur amino acids except for GSH, CysGly, and γ-GluCys, the second containing Cys, CysGly, and γ-GluCys, and the third containing nonsulfur protein amino acids; are prepared in 0.05 M HCl. 2. Working standard solutions: each working standard solution is made up freshly as required by dilution of the stock solution with 0.01 M HCl. When these standard solutions are stored at 4°C, they are stable for at least 2 wk. 3. The GSH standard solution: GSH is freshly dissolved in distilled water at the required concentration on each day, because GSH is partialy hydrolyzed to γGluCys during storage. 4. The internal standard solutions: AE-Cys and LTH are dissolved in 0.05 M HCl to make a stock solution at a concentration of 2 mM. Store at 4°C. 5. NaBH4 solution: NaBH4 is dissolved in 0.1 M NaOH at a concentration of 100 mg/mL solution. This solution is stable at 4°C for at least 1 wk. 6. DTE solution: DTE is dissolved in distilled water at a concentration of 0.5 mM.

3. Methods 3.1. Derivatization Derivatization procedure of sulfur amino acids, GSH and related aminothiols is shown in Fig. 2 (see Note 1). 1. For the analysis of 13 sulfur amino acids, pipet an aliquot of the sample containing 0.5–10 nmol of each amino acid and 0.1 mL of 2 µM AE-Cys (IS) (if necessary) into a 10-mL Pyrex glass reaction tube with a PTFE-lined screw cap.


211

Fig. 2. Shematic flow diagram of the derivatization procedure. 2. For the analysis of GSH and related aminothiols, pipet an aliquot of the sample containing 0.2–25 nmol of each compounds and 0.1 mL of 10 µM LTH (IS) (if necessary) into a reaction tube. 3. To these solutions, add 40 µL of 0.5 mM DTE (see Note 2) and 0.05 mL of 2 M NaOH, and make the total volume up to 1 mL with distilled water. 4. Immediately add 25–50 µL of isoPCF and shake the mixture at 300 cycles per minute (up and down) for 5 min at room temperature (see Note 3). 5. Extract the reaction mixture with 3 mL of peroxide-free diethyl ether (see Note 2), to remove the excess of reagent, with vigorous shaking for 5–10 s by hand (see Note 4). 6. Centrifuge at 2000g for 30 s then discard the etheral extract. 7. Acidify the aqueous layer to pH 1–2 with 2 M HCl and saturate with NaCl. 8. Extract the mixture twice with 3 mL of peroxide-free diethyl ether with vigorous shaking for 5–10 s by hand (see Note 5). 9. Repeat centrifugation at 2000g for 30 s, and transfer the ether layers into another tube by means of a Pasteur capillary pipet. 10. Evaporate the ethereal extracts to dryness at 80°C, and to the residue add 0.2 mL of 1 M HCl-MeOH. Incubate the mixture at 80°C for 10 min (see Note 6). 11. Evaporate the residual solvent to dryness at 80°C under a stream of dry air. Redissolve the residue on the walls of the tube in 0.1–0.2 mL of ethyl acetate and inject 0.2–1 µL of this solution into the gas chromatograph (see Note 7).

212

Kataoka et al.

3.2. Gas Chromatography 1. GC analysis was carried out with a Shimadzu 12A gas chromatograph equipped with a flame ionization detector and a flame photometric detector (S-filter) (see Note 8). 2. For the analysis of sulfur amino acids, a fused-silica capillary column (15 m × 0.53-mm id, 1.0-µm film thickness) of crosslinked DB-17 is used. The operating conditions are as follows: column temperature program at 5°C/min from 130°C to 270°C or at 10°C/min from 150°C to 270°C; injection and detector temperature, 280°C; nitrogen flow rate, 10 mL/min (see Note 9). 3. For the analysis of GSH and related aminothiols, a fused-silica capillary column (5 m × 0.53-mm id, 1.5-µm film thickness) of crosslinked DB-1 is used. The operating conditions are as follows: column temperature program at 16°C/min from 130°C to 290°C; injection and detector temperature, 300°C; nitrogen flow rate, 12 mL/min (see Notes 10 and 11). 4. Measure the peak heights of each compound and the IS. Calculate the peak height ratios against the IS to construct calibration curves (see Notes 12–15).

3.3. Sample Preparation (see Note 16) 3.3.1. Preparation of Urine Samples (44) 1. Collect early morning urine samples from healthy volunteers and processed immediately or store at –20°C until used. 2. Urine samples (0.02–0.1 mL) are used directly for derivatization and FPD-GC analysis (see Note 17).

3.3.2. Preparation of Blood Samples (46) 1. Collect venous blood samples from healthy volunteers in 9 vol of 1 mM ethylenediamine tetraacetate (EDTA) and mix (see Note 18). 2. Immediately process the 10% (v/v) clear hemolysates or store at –20°C until used. 3. To determine the free GSH, Cys, CysGly, and γ-GluCys contents, 0.1 mL of 10% hemolysate is directly used as the sample for derivatization. 4. To obtain total (free and bound, reduced, and oxidized) GSH, Cys, CysGly, and γ-GluCys, chemically reduce a blood sample with NaBH4 (see Note 19). To 0.05 mL of 10% hemolysate add 0.1 mL of 10 µM LTH (IS) and 0.2 mL of 100 mg/mL NaBH4, and then make the total volume up to 1 mL with distilled water. After adding 1 drop of n-hexanol (see Note 20), the mixture is incubated at 100°C for 10 min (see Note 21). After cooling, the reaction mixture is used as the sample for derivatization.

3.3.3. Preparation of Tissue Samples (47) The outline of the method for determining the different forms of GSH and related aminothiols in tissue samples is shown in Fig. 3 (see Note 22). Five male ddY mice 6 wk old (28.0 ± 0.9 g) were used in the experiments.


213

Fig. 3. Outline of method for the determination of different forms of glutathione and related aminothiols in tissue samples.

1. Immediately after dissection, each organ should be removed and stored at –20°C until used. 2. Homogenize an aliquot of each pooled tissue (0.2–0.4 g) with 2.4 mL of 0.05 M H2SO4 and 1.2 mL of 2.5% Na2WO4 with an LK–21 ultradisperser (Yamato Kagaku, Tokyo, Japan) after adding one drop of n-hexanol. 3. Centrifuge at 2000g for 5 min, then reextract the precipitate with the same volume of 0.05 M H2SO4 and 2.5% Na2WO4 as above. 4. Combine the supernatants and make up to 10 mL with distilled water. This solution is then ready to use to determine the free and free oxidized forms. 5. For the analysis of free GSH and other aminothiols, 0.1 mL of a free fraction is directly derivatized. 6. For the analysis of total free GSH and other aminothiols (thiols + disulfides + mixed disulfides with other low-molecular-weight thiols), add 0.1 mL of the free fraction to 0.2 mL of 100 mg/mL of NaBH4 in a 0.01 M NaOH solution, and incubate the mixture at 60°C for 5 min. After cooling, the reaction mixture is used as the sample for derivatization. 7. On the other hand, the precipitate is dissolved in 5 mL of 0.01 M NaOH and made up to 10 mL with distilled water, before being used to determine the proteinbound form. For the analysis of protein-bound GSH and other aminothiols, add 0.5 mL of the protein-bound fraction to 0.2 mL of 100 mg/mL of NaBH4 in a 0.01 M NaOH solution, and incubate the mixture at 60°C for 5 min. After cooling, the reaction mixture is used as the sample for derivatization.

214

Kataoka et al.

4. Notes 1. IsoPCF and HCl-MeOH were used as derivatizing reagents for the amino and sulfhydryl groups, and for the carboxyl group, respectively (Fig. 1). 2. In order to prevent the oxidation of sulfur function during derivatization, DTE was added to reaction mixture and peroxide-free diethyl ether was used as an extraction solvent. 3. The N(S)-isopropoxycarbonylation of sulfur amino acids, GSH, and related aminothiols with isoPCF proceeded rapidly and quantitatively in aqueous alkaline media (44,46). This reaction was completed within 5 min by shaking at room temperature. 4. The reaction mixture was then washed with diethyl ether under alkaline condition in order to remove the excess reagent. This procedure also serves to exclude amines and phenols both of which often coexist with amino acids in biological samples, as they are derivatized to the corresponding N- or O-isoPOC derivatives, which are soluble in organic solvents under the same conditions as aforementioned. 5. The resulting N(S)-isoPOC derivatives in aqueous layer were quantitatively and selectively extracted into diethyl ether after acidification to pH 1.0–2.0. In this procedure, the ether layers should be collected, taking care to avoid aqueous droplets. It was not necessary to complete draw the ether layer in each extraction. 6. Although the methylation of the carboxyl group with diazomethane was simple (53,54), the interfering peaks originating in the diazomethane generating reagents were observed under our FPD-GC conditions. The methylation of N(S)-isoPOC derivatives with HCl-MeOH was completed within 5 min at 80°C, and no interfering peak was observed. 7. The derivative preparation was accomplished within 30 min, and several samples could be treated simultaneously. The N(S)-isoPOC methyl ester derivatives of sulfur amino acids and related compounds were stable under nomal laboratory conditions and no decomposition was observed during GC analysis. 8. The GC analysis was performed with FPD (394-nm interference filter), which is highly selective for sulfur containing compounds, by using a megabore capillary column and direct (splitless) sample injection system. In this system, the sufficiently inactivated glass insert should be used to avoid tailing peaks of derivatives caused by the adsorption on injection port. 9. Nonsulfur amino acids were also derivatized and detected with flame ionization detection (FID) (Fig. 4A), but these amino acids were not detected at all with FPD (Fig. 4C). On the other hand, sulfur amino acids were detected with both FID (Fig. 4B) and FPD (Fig. 4D). As shown in Fig. 4, 13 sulfur amino acids and IS could be completely resolved as single and symmeterical peaks within 32 min on a DB-17 capillary column (15 m × 0.53-mm id, 1.0-µm film thickness) (44). 10. The separation of biologically important sulfur amino acids was achieved within 18 min on a DB-17 capillary column (Fig. 5A). However, the derivatives of GSH and γ-GluCys could not be eluted with this capillary column because of high molecular mass and high boiling point of these derivatives.


215

Fig. 4. Standard chromatograms of the N(O,S)-isoprpoxycarbonyl methyl ester derivatives of nonsulfur and sulfur amino acids. (A) Nonsulfur amino acids (containing 200 nmol of each amino acid); (B) sulfur amino acids (containing 200 nmol of each amino acid); (C) nonsulfur amino acids (containing 50 nmol of each amino acid); (D) sulfur amino acids (containing 1 nmol of each amino acid). The derivatized samples were analyzed by FID-GC (A and B) and FPD-GC (C and D). GC conditions: column, DB-17 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, programmed at 5°C/min from 130°C to 270°C and then held for 5 min; injection and detector temperatures, 280°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = alanine; 2 = glycine; 3 = valine; 4 = leucine; 5 = isoleucine; 6 = serine + threonine + proline; 7 = aspartic acid; 8 = glutamic acid; 9 = hydroxyproline; 10 = phenylalanine; 11 = asparagine; 12 = glutamine; 13 = lysine; 14 = histidine; 15 = tyrosine; 16 = tryptophan; 17 = S-methylcysteine (SM-Cys); 18 = thioproline (TPro); 19 = methionine (Met); 20 = ethionine (Eth); 21 = cysteine (Cys); 22 = S-carboxymethylcysteine (SC-Cys); 23 = homocysteine (HCys); 24 = methionine sulfone (Met-S); 25 = S-2-aminoethylcysteine (AE-Cys) (IS); 26 = lanthionine (LTH); 27 = cystathionine (CTH); 28 = cystine (Cyt); 29 = djenkolic acid (Dje); 30 = homocystine (HCyt) (from ref. 44, with permission).

216

Kataoka et al.

Fig. 5. Chromatograms obtained from standard and urine samples. (A) Standard (containing 2 nmol of each amino acid); (B) urine A (20 µL); (C) urine B (20 µL); (D) urine B (0.8 mL). The derivatized samples were analyzed by FPD-GC (A–C) and FIDGC (D). GC conditions: column, DB-17 (15 m × 0.53-mm id, 1.0-µm film thickness); column temperature, programmed at 10°C/min from 150°C to 270°C and then held for 5 min; injection and detector temperatures, 280°C; nitrogen carrier gas flow rate, 10 mL/min. Peaks: 1 = methionine (Met); 2 = cysteine (Cys); 3 = homocysteine (HCys); 4 = S-2-aminoethylcysteine (AE-Cys) (IS); 5 = cystathionine (CTH); 6 = cystine (Cyt); 7 = homocystine (HCyt) (from ref. 44, with permission).

11. In order to solve above problem, a column was cut short. Of different length of several columns tested, DB-1 (5 m × 0.53-mm id, 1.5-µm film thickness) proved to be the most satisfactory column for the purpose. By using this column, Cys, CysGly, γ-GluCys, GSH, and IS were separately eluted as single and symmetrical peaks within 10 min (Fig. 6A) (46). However, GSSG, disulfide form of GSH, could not be eluted even with this column. 12. The calibration curves for sulfur amino acids by FPD-GC were conducted using AE-Cys, which showed a similar behavior to other sulfur amino acids during the derivatization and was well separated from these amino acids on a chromatogram as the IS (44,45). A linear relationship was obtained from double logarithmic plots in the range 0.5–10 nmol of sulfur amino acids, and correlation coefficients were above 0.998 (Table 1). 13. The derivatives of sulfur amino acids provided excellent FPD responses and the minimum detectable amounts of these amino acids at a signal three times as high as the noise under our instrumental conditions were approx 0.5–1.0 pmol as injec-


217

Fig. 6. Chromatograms obtained from standard and whole blood samples. (A) Standard (containing 5 nmol of glutathione and 1 nmol of other aminothiols); (B) blood (10 mL, nonreduction); (C) blood (5 mL, reduction). GC conditions: column, DB-1 (5 m × 0.53-mm id, 1.5-µm film thickness); column temperature, programmed at 16°C/ min from 130°C to 290°C; injection and detection temperature, 300°C; nitrogen flow rate, 12 mL/min. Peaks: 1 = cysteine (Cys), 2 = cysteinylglycine (CysGly), 3 = lanthionine (LTH) (IS), 4 = γ-glutamylcysteine (γ-GluCys), 5 = glutathione (GSH) (from ref. 46, with permission).

tion amounts. The FPD-GC system described here was over 20 times more sensitive than the FID-GC system. 14. The calibration curves for GSH and related aminothiols by FPD-GC were conducted using LTH, which showed a similar behavior to other aminothiols during the derivatization and was well separated from these aminothiols on a chromatogram as the IS (46,47). A linear relationship was obtained from double-logarithmic plots in the range 1–25 nmol for GSH and 0.2–5 nmol for other aminothiols, and correlation coefficients were above 0.999 (Table 2). 15. The minimum detectable amount of GHS and related aminothiols to give a signal three times as high as the noise under our FPD-GC conditions were approx 0.2– 5 pmol as injection amounts.

218

Table 1 Linear Regression, Detection Limits, and Recoveries from Urine Samples on the Determination of Sulfur Amino Acids Sulfur amino acid Met Cys HCys CTH Cyt HCyt

Regression linea Detection Urine A Urine B Correlation limit Amount found (nmol/mL)b Recovery Amount found (nmol/mL)b Recovery Slope Intercept coefficient (pmol) Non-addition Additionc (%) Non-addition Additionc (%) 1.822 1.777 1.703 1.781 1.834 1.859

–0.222 –0.308 –0.514 –0.548 –0.188 –0.059

0.9983 0.9989 0.9995 0.9998 0.9998 0.9986

0.5 0.5 0.5 1.0 1.0 1.0

12.9 ± 0.6 33.9 ± 1.2 NDd 22.1 ± 0.9 45.0 ± 1.2 ND

111.0 ± 6.2 136.5 ± 1.3 90.0 ± 2.2 122.0 ± 1.2 145.0 ± 0.9 113.0 ± 1.3

98.1 102.6 90.9 99.9 100.0 113.0

8.0 ± 0.5 69.0 ± 1.5 ND 22.0 ± 0.3 61.5 ± 0.3 ND

93.0 ± 0.3 165.0 ± 4.5 92.5 ± 2.5 114.5 ± 2.5 159.0 ± 3.5 107.5 ± 1.5

85.0 96.0 92.5 92.5 97.5 107.5

alog

y = a log x + b : y, peak height ratio against the IS; x, amount of each amino acid (nmol); a, slope; b, intercept. Range: 0.5–10 nmol. ± SD (n = 3). cAddition: 100 nmol/mL. dNot detectable. bMean

Kataoka et al.

Regression linea Detection Non-reduced blood Reduced blood b Correlation limit Amount found (nmol/mL) Recovery Amount found (nmol/mL)b Recovery Aminothiol Slope Intercept coefficient (pmol) Non-addition Additionc (%) Non-addition Additionc (%) Cys CysGly γ-GluCys GSH

1.418 1.419 1.419 1.677

0.367 0.186 –0.029 –1.154

0.9997 0.9992 0.9990 0.9993

0.2 0.3 0.5 5.0

19.0 ± 1.4 19.5 ± 0.9 14.3 ± 1.0 464 ± 34

111.2 ± 3.8 110.2 ± 8.3 104.6 ± 3.2 954.0 ± 42

92.2 90.7 90.3 98.0

83.3 ± 2.2 55.4 ± 2.6 72.6 ± 4.8 958 ± 55

260.7 ± 8.3 248.9 ± 13.1 286.0 ± 7.7 1977 ± 121

88.7 96.8 106.7 101.9


Table 2 Linear Regression, Detection Limits, and Recoveries from Blood Samples on the Determination of GSH and Related Aminothiols

alog y = a log x + b : y, peak height ratio against the IS; x, amount of each amino acid (nmol); a, slope; b, intercept. Range: 1–25 nmol for GSH, 0.2–5 nmol for the other aminothiols. bMean ± SD (n = 3). cAddition: 500 nmol/mL for GSH, 100 nmol/mL for the other aminothiols. dAddition: 1000 nmol/mL for GSH, 200 nmol/mL for the other aminothiols.

219

220

Kataoka et al.

16. The method developed for sulfur amino acids was successfully applied to small urine (0.02–0.1 mL) (44) and plasma (0.1 mL) (45) samples without prior cleanup of the sample. As shown in Fig. 5B,C, unknown peaks were observed in front of the Cys peak and in the near of the IS peak, but Met, Cys, CTH, Cyt, and trace quantities of HCys and HCyt were detected without any interference from coexisting substances when the sample was analyzed by FPD-GC. 17. It was difficult to determine sulfur amino acids in urine by FID-GC (Fig. 5D). As shown in Table 1, the overall recoveries of sulfur amino acids added to urine samples were 85–113% and the relative standard deviations were 0.3–6.4% (n = 3). 18. The method developed for GSH and related aminothiols was successfully applied to blood sample, in which whole blood was used because it should reflect more accurately the total GSH status of blood including GSH bound to plasma proteins, and erythrocytes contain almost all of the GSH (55). Hemolysis of red blood cell was carried out in solution of EDTA to prevent oxidation of thiol function by any metallic elements. The hemolysate of small blood (5–10 µL) could be used without prior clean-up of the sample (46). Figure 6B,C show typical chromatograms obtained from nonreduced and reduced blood samples, respectively. Free and total GSH and related aminothiols in whole blood sample could be analyzed without any interference from coexisting substances. As shown in Table 2, the overall recoveries of GSH and related aminothiols added to whole blood samples were 88–107% and the relative standard deviations were 2.6–7.5% (n = 4). 19. In order to determine the redox status of sulfur amino acids (45), GSH, and related aminothiols (46,47), disulfides in the sample must be reduced to the corresponding thiols prior to derivatization. NaBH4 efficiently converts all these aminothiols present as free thiols and mixed disulfides with other low-molecular mass thiols or protein-thiols to thiol forms (20–22). 20. Formation of gas and foaming during the NaBH4 reduction of biological samples could be reduced by adding one drop of n-hexanol, surface active agent. The reduced sample could be directly derivatized after cooling. 21. In order to confirm the reduction conditions, some disulfide compounds such as Cyt, HCyt, and GSSG were reduced with NaBH4. The reduction of these compounds were accomplished within 10 min at 100°C by using 20 mg NaBH4 in aqueous alkaline medium. 22. In order to determine the concentration of the different forms of GSH and related aminothiols in tissue samples, the fractionation of the tissue samples was investigated according to the previous method (56). As shown in Fig. 3, the sample was separated into its free and protein-bound fractions by precipitation with Na2WO4H2SO4. No free GSH or other aminothiols were apparent in the protein-bound fraction. Figure 7B–D shows typical chromatograms obtained from the free fraction (-reduction), free fraction (+reduction), and protein-bound fraction (+reduction), respectively. Each form of GSH and related aminothiols in a tissue sample could be analyzed without any interference from coexisting substances. The overall recoveries of GSH and related aminothiols added to the free and protein-bound


221

Fig. 7. Gas chromatograms obtained from standard solution and free and proteinbound fractions of mouse liver. (A) Standard containing 0.25 nmol of cysteine, 0.5 nmol of cysteinylglycine and γ-glutamylcysteine and 5 nmol of glutathione; (B) free fraction (–reduction); (C) free fraction (+reduction); (D) protein-bound fraction (+reduction). GC conditions and peak no.: see Fig. 6. The double arrow on the time-scale indicates the time during which the recorder response was reduced to half (from ref. 47 with permission). fractions of mouse liver were 91.2–109.6%, and the relative standard deviation was 1.1–10.7% (n = 4). Oxidized GSH and related aminothiols (disulfides and mixed disulfides with low-molecular weight thiols) were measured by subtracting their free reduced contents from their total free contents.

References 1. Mizobuchi, N., Ageta, T., Sasaki, K., and Kodama, H. (1986) Isotacho- phoretic analysis of cystine, homocystine and cystathionine in urines from patients with inborn errors of metabolism. J. Chromatog. 382, 321–325. 2. Friedman, M., Noma, A. T., and Wagner, J. R. (1979) Ion-exchange chromatography of sulfur amino acids on a single-column amino acid analyzer. Anal. Biochem. 98, 293–304.

222

Kataoka et al.

3. Andersson, A., Brattstrom, L., Isaksson, A., Israelsson, B., and Hultberg, B. (1989) Determination of homocysteine in plasma by ion-exchange chromatography. J. Clin. Lab. Invest. 49, 445–449. 4. Caldwell, K. A. and Tappel, A. L. (1968) Separation by gas-liquid chromatography of silylated derivatives of some sulfo- and selenoamino acids and their oxidation products. J. Chromatog. 32, 635–640. 5. Shahrokhi, F. and Gehrke, C. W. (1968) Quantitative gas-liquid chromatography of sulfur containing amino acids. J. Chromatog. 36, 31–41. 6. Jellum, E., Bacon, V. A., Patton, W., Pereira, W., Jr., and Halpern, B. (1969) Quantitative determination of biologically important thiols and disulfides by gas-liquid chromatography. Anal. Biochem. 31, 339–347. 7. Mee, J. M. L. (1973) Gas-liquid chromatographic analysis of glutathione. J. Chromatog. 87, 258–262. 8. Bonvel, S. I. and Monheimer, R. H. (1980) A gas chromatographic analysis of sulfur-containing amino acids employing flame photometric detection. J. Chromatog. Sci. 18, 18–22. 9. Reed, D. J., Babson, J. R., Beatty, P. W., Brodie, A. E., Ellis, W. W., and Potter, D. W. (1980) High-performance liquid chromatography analysis of nanomole levels of glutathione, glutathione disulfide, and related thiols and disulfides. Anal. Biochem. 106, 55–62. 10. Di Pietra, A. M., Gotti, R., Bonazzi, D., Andrisano, V., and Cavrini, V. (1994) HPLC determination of glutathione and L-cysteine in pharmaceuticals after derivatization with ethacrynic acid. J. Pharm. Biomed. Anal. 12, 91–98. 11. Asensi, M., Sastre, J., Pallardo, F. V., de la Asuncion, J. G., Estrela, J. M., and Vina, J. (1994) A high-performance liquid chromatography method for measurement of oxidized glutathione in biological samples. Anal. Biochem. 217, 323–328. 12. Yoshida, T. (1996) Determination of reduced and oxidized glutathione in erythrocytes by high-performance liquid chromatography with ultraviolet absorbance detection. J. Chromatog. B 678, 157–164. 13. Andersson, A., Isaksson, A., Brattstrom, L., and Hultberg, B. (1993) Homocysteine and other thiols determined in plasma by HPLC and thiol-specific postcolumn derivatization. Clin. Chem. 39, 1590–1597. 14. Nozal, M. J., Bernal, J. L., Toribio, L., Marinero, P., Moral, O., Manzanas, L., and Rodriguez, E. (1997) Determination of glutathione, cysteine and N-acetylcysteine in rabitt eye tissues using high-performance liquid chromatography and post-column derivatization with 5, 5-dithiobis(2-nitrobenzoic acid). J. Chromatog. A, 778, 347–353. 15. Ubbink, J. B., Vermaak, W. J. H., and Bissbort, S. (1991) Rapid high- performance liquid chromatographic assay for total homocysteine levels in human serum. J. Chromatog. 565, 441–446. 16. Cornwell, P. E., Morgan, S. L., and Vaughn, W. H. (1993) Modification of a highperformance liquid chromatographic method for assay of homocycteine in human plasma. J. Chromatog. 617, 136–139.


223

17. Gotti, R., Andrisano, V., Gotti, R., Cavrini, V., and Candeletti, S. (1994) Determination of glutathione in biological samples by high performance liquid chromatography with fluorescence detection. Biomed. Chromatogr. 8, 306–308. 18. Winters, R. A., Zukowski, J., Ercal, N., Mattews, R. H., and Spitz, D. R. (1995) Analysis of glutathione, glutathione disulfide, cysteine, homocysteine, and other biological thiols by high-performance liquid chromatography following derivatization by N-(1-pyrenyl)maleimide. Anal. Biochem. 227, 14–21. 19. Newton, G. L., Dorian, R., and Fahey, R. C. (1981) Analysis of biological thiols: Derivatization with monobromobimane and separation by reversed-phase high-performance liquid chromatography. Anal. Biochem. 114, 383–387. 20. Jacobsen, D. W., Gatautis, V. J., and Green, R. (1989) Determination of plasma homocysteine by high-performance liquid chromatography with fluorescence detection. Anal. Biochem. 178, 208–214. 21. Mansoor, M. A., Svardal, A. M., and Ueland, P. M. (1992) Determination of the in vivo redox status of cysteine, cysteinylglycine, homocysteine, and glutathione in human plasma. Anal. Biochem. 200, 218–229. 22. Fiskerstrand, T., Refsum, H., Kvalheim, G., and Ueland, P. M. (1993) Honocysteine and other thiols in plasma and urine: Automated determination and sample stability. Clin. Chem. 39, 263–271. 23. Yan, C. C. and Huxtable, R. J. (1995) Fluorimetric determination of monobromobimane and o-phthalaldehyde adducts of γ-glutamylcysteine and glutathione: application to assay of γ-glutamylcysteine synthetase activity and glutathione concentration in liver. J. Chromatog. B 672, 217–224. 24. Fermo, I., Arcelloni, C., De Vecchi, E., Vigano, S., and Paroni, R. (1992) Highperformance liquid chromatographic method with fluorescence detection for the determination of total homocyst(e)ine in plasma. J. Chromatog. 593, 171–176. 25. Paroni, R., DeVecchi, E., Cighetli, G., Arecelloni, C., Fermo, I., Grossi, A., and Bonini, P. (1995) HPLC with o-phthalaldehyde precolumn derivatization to measure total, oxidized, and protein-bound glutathione in blood, plasma, and tissue. Clin. Chem. 41, 448–454. 26. Staffeldt, B., Brockmoller, J., and Roots, I. (1991) Determination of S- carboxymethyl-L-cysteine and some of its metabolites in urine and serum by high-performance liquid chromatography using fluorescence pre-column labelling. J. Chromatog. 571, 133–147. 27. Sypniewski, S. and Bald, E. (1994) Ion-pair high-performance liquid chromatography of cysteine and metabolically related compounds in the form of their Spyridinium derivatives. J. Chromatog. A 676, 321–330. 28. Richie, J. P. and Lang, C. A. (1987) The determination of glutathione, cyst(e)ine, and other thiols and disulfides in biological samples using high-performance liquid chromatography with dual electrochemical detection. Anal. Biochem. 163, 9–15. 29. Carvalho, F. D., Remiao, F., Vale, P., Timbrell, J. A., Bastos, M. L., and Ferreira, M. A. (1994) Glutathione and cysteine measurement in biological samples by HPLC with a glassy carbon working detector. Biomed. Chromatogr. 8, 134–136.

224

Kataoka et al.

30. Yang, C.-S., Tsai, P.-J., Chen, W.-Y., Liu, L., and Kuo, J.-S. (1995) Determination of extracellular glutathione in livers of anaesthetized rats by microdialysis with online high-performance liquid chromatography. J. Chromatog. 667, 41–48. 31. Kleinman, W. A. and Richie, J. P., Jr. (1995) Determination of thiols and disulfides using high-performance liquid chromatography with electrochemical detection. J. Chromatog. B 672, 73–80. 32. Lakritz, J., Plopper, C. G., and Buckpitt, A. R. (1997) Validated high-performance liquid chromatography-electrochemical method for determination of glutathione and glutathione disulfide in small tissue samples. Anal. Biochem. 247, 63–68. 33. Stabler, S. P., Marcell, P. D., Podell, E. R., and Allen, R. H. (1987) Quantitation of total homocysteine, total cysteine, and methionine in normal serum and urine using capillary gas chromatography-mass spectrometry. Anal. Biochem. 162, 185–196. 34. Sass, J. O. and Endres, W. (1997) Quantitation of total homocysteine in human plasma by derivatization to its N(O, S)-propoxycarbonyl propyl ester and gas chromatography-mass spectrometry. J. Chromatog. A 776, 342–347. 35. Yu, S., Sugahara, K., Zhang, J., Ageta, T., Kodama, H., Fontana, M., and Dupre, S. (1997) Simultaneous determination of urinary cystathionine, lanthionine, S-(2aminoethyl)-L-cysteine and their cyclic compounds using liquid chromatographymass spectrometry with atmospheric pressure chemical ionization. J. Chromatog. B 698, 301–307. 36. Lin, B. L., Colon, L. A., and Zare, R. N. (1994) Dual electrochemical detection of cysteine and cystine in capillary zone electrophoresis. J. Chromatog. A 680, 263–270. 37. Zhou, J., O’Shea, T. J., and Lunte, S. M. (1994) Simultaneous detection of thiols and disulfides by capillary electrophoresis-electrochemical detection using a mixedvalence ruthenium cyanide-modified microelectrode. J. Chromatog. A 680, 271– 277. 38. Vesphalec, R., Corstjens, H., Billiet, H. A. H., Frank, J., and Luyben, K. Ch. A. M. (1995) Enantiomeric separation of sulfur- and selenium-containing amino acids by capillary electrophoresis using vancomycin as a chiral selector. Anal. Chem. 67, 3223–3228. 39. Jacoby, W. B. and Griffith, O. W. (eds.) (1987) Sulfur and sulfur amino acids, in Methods in Enzymology, vol. 143, Academic, London and New York. 40. Walker, V. and Mills, G. A. (1995) Quantitative methods for amino acid analysis in biological fluids. Ann. Clin. Biochem. 32, 28–57. 41. Ueland, P. M., Refsum, H., Stabler, S. P., Malinow, M. R., Andersson, A., and Allen, R. H. (1993) Total homocysteine in plasma or serum: methods and clinical applications. Clin. Chem. 39, 1764–1779. 42. Shimada, K. and Mitamura, K. (1994) Derivatization of thiol-containing compounds. J. Chromatog. B 659, 227–241. 43. Zahn, H. and Gattner, H. G. (1997) Hair sulfur amino acid analysis. EXS 78, 239– 258. 44. Kataoka, H., Tanaka, H., Fujimoto, A., Noguchi, I., and Makita, M. (1994) Determination of sulphur amino acids by gas chromatography with flame photometric detection. Biomed. Chromatogr. 8, 119–124.


225

45. Kataoka, H., Takagi, K., and Makita, M. (1995) Determination of total plasma homocysteine and related aminothiols by gas chromatography with flame photometric detection. J. Chromatog. B 664, 421–425. 46. Kataoka, H., Takagi, K., and Makita, M. (1995) Determination of glutathione and related aminothiols by gas chromatography with flame photometric detection. Biomed. Chromatogr. 9, 85–89. 47. Takagi, K., Kataoka, H., and Makita, M. (1996) Determination of glutathione and related aminothiols in mouse tissues by gas chromatography with flame photometric detection. Biosci. Biotech. Biochem. 60, 729–731. 48. Kataoka, H., Yamamoto, S., and Makita, M. (1984) Quantitative gas-liquid chromatography of taurine. J. Chromatog. 306, 61–68. 49. Kataoka, H., Ohnishi, N., and Makita, M. (1985) Electron-capture gas chromatography of taurine as its N-pentafluorobenzoyl di-n-butylamide derivative. J. Chromatog. 339, 370–374. 50. Kataoka, H., Ohishi, K., and Makita, M. (1986) Gas chromatographic determination of cysteic acid. J. Chromatog. 354, 482–485. 51. Kataoka, H., Yamamoto, H., Sumida, Y., Hashimoto, T., and Makita, M. (1986) Gas chromatographic determination of hypotaurine. J. Chromatog. 382, 242–246. 52. Okazaki, T., Kataoka, H., Fujimoto, A., Kono, K., and Makita, M. (1989) Determination of taurine in biological samples by GC with flame photometric detection. Bunseki Kagaku (Japanese) 38, 401–403. 53. Makita, M., Yamamoto, S., and Kono, M. (1976) Gas-liquid chromatographic analysis of protein amino acids as N-isobutoxycarbonylamino acid methyl esters. J. Chromatog. 120, 129–140. 54. Yamamoto, S., Kiyama, S., Watanabe, Y., and Makita, M. (1982) Practical gasliquid chromatographic method for the determination of amino acids in human serum. J. Chromatog. 233, 39–50. 55. Mills, B. J., Richie, J. P., Jr., and Lang, C. A. (1990) Sample processing alters glutathione and cysteine values in blood. Anal. Biochem. 184, 263–267. 56. Kataoka, H., Imamura, Y., Tanaka, H., and Makita, M. (1993) Determination of cysteamine and cystamine by gas chromatography with flame photometric detection. J. Pharm. Biomed. Anal. 10, 963–969.

CE Determination of 4-Hyp

227

16 Capillary Electrophoretic Determination of 4-Hydroxyproline Qingyi Chu and Michael Zeece 1. Introduction Capillary electrophoresis (CE) represents a relatively new separation technology that has gained acceptance in a wide variety of applications. A discussion of the basic theory of CE separations is beyond the scope of this presentation, but is well addressed in a number of recent texts (1–5). In principle, differential migration of analytes in a potential field is achieved as a result of individual differences in mass/charge ratios. In CE, substantial advantage in separation speed, efficiency, and resolution is derived from the veryhigh field strengths (e.g., 150–300 V/cm) typically used. In addition, high field strengths induce the effect of electroosmosis in the column. The walls of the silica column have a negative charge and attract hydrated counter ions from the buffer. When power is applied to the system, the positively charged ions with their associated water molecules, migrate toward the cathode with substantial velocity. This results in the flow of water termed electroosmosis. The flow proceeds from the anode to the cathode and serves as a pump. Because the electroosmotic water flow is much greater than the velocity of the analytes, all components are swept to the cathode. Most columns used in CE have an internal diameter of 50–100 µm and can be 1 m long. High field potentials can be used because the column has a large surface area to internal volume ratio that facilitates rapid dissipation of heat. Extremely small volume samples are required for CE analysis. For example, a typical amino acid separation, such as those presented here, employ 10–20 nL sample injections. The small volume of sample injection however, also presents a challenge for detection. In order to achieve good sensitivities with UV detection, it is necessary to label the amino acids with phenylisothiocynate From: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

227

228

Chu and Zeece

(PITC) prior to analysis. This reaction converts the amino acids to their phenylthiohydantoin (PTH) derivatives. The mode of CE separation used for the amino acid analysis described here employs the anionic detergent sodium dodecyl sulfate (SDS). CE performed in the presence of SDS is often termed a micellar electrokinetic chromatography (MEKC) separation (6). When SDS concentrations are above 8–10 mM, the molecules form micelles in which the hydrophobic tails are oriented toward the center and sulfate groups are on the surface. The micelles have a strong negative charge at the pHs employed in most MEKC separations and migrate toward the anode. Their velocity, however, is less than that of the electroosmotic flow and all components are carried to the cathode. The SDS micelle thus becomes a pseudostationary phase with which analytes can interact. The resulting partitioning provides additional selectivity and the technique has been extensively used for the separation and determination of amino acids and other compounds (7). 2. Materials 2.1. Equipment 1. CE units are available from several well-known vendors and the method described here was performed with an ISCO (Lincoln, NE) model 3850 capillary electropherograph. Samples are loaded into the capillary by vacuum injection and detection of separated analytes is performed at 254 nm. 2. A CAESAR (Version 4.01, 1994, Roman Scientific) from Scientific Resources (Eatontown, NJ) was used to record and analyze data. 3. A Centri-Vap (Fisher Scientific, Pittsburgh, PA) or similar device for drying small volumes in microfuge tubes. 4. An amine coated capillary (50-µm id Polymicro Technology, Phoenix AZ).

2.2. Reagents Highest purity reagents were used throughout the experiments. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Phenyl isothiocyanate was obtained from Pierce (Rockford, IL). Sodium hydroxide. Hydrochloric acid. Triethylamine. Diethanolamine. Malonic acid . Chloro(chloro.methyl)dimethylsilane. Phthalic dicarboxaldehyde. 3- and 4-Hyp. Cis and trans 4-Hyp. Methanol. Sodium phosphate (monobasic).


229

13. SDS. 14. Thymine (as internal standard). 15. Water used in these experiments was obtained by reverse osmosis (18 mOhm).

2.3. Amine Capillary Preparation An amine coated capillary was used in these separations (50-µm id Polymicro Technology, Phoenix, AZ). The length to the detector was 40 cm and total column length was 70 cm. This column can be purchased from commercial vendors or can be made in the lab (see Note 1). The procedure for coating the capillary with diethanolamine was originally described by Kuhn and Hoffstetter-Kuhn (8) and is summarized as follows. 1. Using a 1-mL syringe fitted with an adapter, fill 1 m of fused silica capillary with 1 M NaOH and let stand at room temperature (20°C) for 1 h. 2. Rinse the capillary twice with water. 3. Rinse the capillary twice with methanol followed by two rinses with acetone 4. Fill the capillary with 10% (w/v) chloro(chloromethyl)-dimethylsilane in methanol and hold for 30 min at room temperature. 5. Cover the ends of the silane-filled capillary with tape and incubate at 100°C for 3 h. 6. Flush the column with first with methanol, then with water. 7. Fill the capillary with 3.0 M diethanolamine HCl dissolved in methanol, and incubate overnight at room temperature.

3. Methods

3.1. PTH Labeling of Hyp Standards Standards are necessary for identification of the Hyp peak in the separation and for development of standard curves. The preparation of these standards as PTH derivatives of Hyp isomers was performed essentially as described by Chu et al. (9) and summarized below. 1. Dissolve 6.6 mg of Hyp in 1.0 mL of coupling buffer. This solution contains methanol-triethylamine-water in 7:1:1 (v/v) ratio. A 1.5 or 2.0 mL microfuge tube works well for holding this mixture. 2. Flush the tube with nitrogen gas and add 10 µL of PITC per 100 µL of sample. 3. Incubate the reaction for 10 min at room temperature and then dry in a CentriVap (Fisher Scientific). 4. The samples are stored dry at –70°C and are dissolved in 100 µL acetonitrile just prior to electrophoresis (see Note 2).

3.2. Preparation of Muscle Samples Hyp determination can be made from isolated collagen fractions or whole muscle tissues samples. The procedures for preparing these materials for labeling is summarized as follows (see Note 3).

230

Chu and Zeece

3.2.1. Isolated Collagen Fractions Perimysial or other types of collagen can be isolated from skeletal muscle tissue using the well-established procedures described by Light and Champion (10). Following isolation, the collagen samples are prepared as follows (see Note 4). 1. Desiccate the material by washing three times in cold (4°C) acetone. For each wash, the sample is suspended in 4 vol of cold acetone, stirred by hand and then centrifuged at 2000g for 10 min. The supernatant is discarded each time. Be sure to use polyethylene type tubes, because clear plastics such as polycarbonate, are attacked by the solvent. 2. Dry the sample in a hood under a stream of nitrogen. 3. Add 50 mg of collagen to a suitable hydrolysis vial. 4. Add 5.0 mL of 6 M HCl, draw a vacuum, and heat for 24 h at 110°C. 5. Centrifuge or filter the sample after hydrolysis to remove any insolubles. 6. Withdraw 100-µL aliquots (equivalent to approx 1.0 mg collagen sample) and place in a microfuge tube. 7. Dry the samples in a Centri-Vap. The dried samples are now ready for labeling, but can also be stored at –70°C.

3.2.2. Whole Muscle To analyze whole muscle tissue for Hyp content, the tissue must first be dried and then hydrolyzed. The preparation of muscle samples is as follows (see Note 4). 1. Grind or finely dice 10 g of muscle. 2. Add 100–200 mg of ground muscle to a mortar and pestle prechilled with liquid nitrogen. Add liquid nitrogen slowly and grind to a powder. Keep the sample frozen hard during grinding with subsequent additions of liquid nitrogen. 3. Lyophilize the sample. 4. Add 65 mg of dried muscle powder in a suitable hydrolysis vial. 5. Add 5 mL of 6 M HCl, draw a vacuum and heat for 24 h at 110°C. 6. Centrifuge or filter the sample after hydrolysis to remove any insolubles. 7. Withdraw 150-µL aliquots (equivalent to about 2.0 mg muscle) and place in a microfuge tube. 8. Dry the samples in a Centri-Vap. The dried samples are now ready for labeling, but can also be stored at –70°C.

3.3. Labeling of Amino Acids from Hydrolyzed Samples The strategy for labeling amino acids from hydrolyzed samples is based on a modification of the procedure described by Yaegaki et al. (11). Amino acids from hydrolyzed tissues are labeled using a two-step procedure that involves initial reaction with o-phthalaldehyde (OPA) to react with primary amines fol-


231

lowed by their precipitation under acidic conditions. In the second step, amino acids (Pro and Hyp) are reacted with PITC. The procedure is summarized as follows (see Note 5). 1. Add 100 µL of 0.5 M Na2 CO3, NaHCO3 pH 9.5, to the dried aliquot of hydrolyzed sample. The sample should contain approx 1.0 mg of hydrolyzed material. 2. Add 100 µL (2 mg) of 20 mg/mL OPA reagent, mix, and let stand at room temperature for 15 min. 3. Precipitate the reaction products by adding of 900 µL of 1.0 M NaH2PO4 pH 2.5. Be sure to mix the sample well (e.g., in a vortex mixer). 4. Centrifuge at 10,000g for 5 min and the filter the supernatant through 0.45-µm spin-filters (Millipore, Bedford, MA) 5. Dry 100 µL of filtrate in the Centri-Vap. 6. Add 150 µL coupling buffer (described in Subheading 3.1.). 7. Flush the vial with nitrogen and add 10 µL of PITC. 8. Incubate the mixture at room temperature for 10 min and Centri-Vap to dryness. 9. The dried samples are ready for CE analysis or can be stored at –70°C.

3.4. CE Analysis Samples are ready for analysis after the appropriate preparation steps have been performed. The procedure for CE for identification and determination of Hyp in these samples is described below (see Note 6). 1. Add 100 µL 50% acetonitrile to the tube containing the dried and labeled amino acid sample (prepared in Subheading 3.3.). Mix well by vortexing. 2. Centrifuge at 10,000g for 2 min. 3. Withdraw 30 µL of the sample supernatant and add 10 µL 2.0 mM thymine (internal standard). 4. Equilibrate the column filling it with the run buffer (50 mM Na-malonate, 75 mM SDS, pH 5.0). 5. The sample is then loaded to the CE using either vacuum or pressure injection following the CE manufacturer’s instructions. Generally, loads will be in the order of 5 to 50 nL. The equipment manufacturer should provide information regarding determination of load volumes. 6. Set the instrument voltage at 15 kV and start the run. 7. Record the data for all peaks. 8. Identify the thymine and Hyp peaks in the separation from the procedure described below.

3.5. Quantitation of Hyp in Samples In order to determine the amount of Hyp accurately in unknown samples, it is first necessary to identify the peak in the mixture using coinjection and then use an internal standard method for its quantitation. The procedures described here were developed for determination of Hyp in muscle tissue samples (9). It

232

Chu and Zeece

Fig. 1. Electropherogram of PTH-Proline, 3- and 4-Hydroxyproline standards. Amino acids were labeled with PITC as described in Subheading 3.1. to obtain the PTH derivatives. Approximately 20 nL of a mixture containing 2.0 mM each of PTHProline, 3- and 4-Hydroxyproline was separated by capillary electrophoresis in 50 mM Na-malonate, 75 mM SDS, pH 5.0 at 15 kV. The capillary was amine coated, 50 µm in id and 70 cm in length (40 cm to the detector). The migration of the components are identified in the figure.

is important for accurate quantitation that the Hyp peak be reasonably separated from other peaks in the sample. Thus it may be necessary to adjust parameters such as voltage or buffer composition when analyzing samples prepared from other tissues.

3.5.1. Identification of Hyp in the Separation (Electropherogram) In order to identify the Hyp peak in a sample, the migration time must first be determined using standards (prepared in Subheading 3.1.) (see Note 7). A sample electropherogram illustrating the separation of proline, 3-hydroxyproline and 4-hydroxyproline is shown in Fig. 1. The procedure is as follows. 1. Dilute the appropriate PTH-Hyp isomer to 2.0 mM in acetonitrile. 2. Add 10 µL of 2.0 mM thymine per 30 µL of Hyp solution. 3. Mix and perform CE separation as described in Subheading 3.4. The conditions for separation of the standard (s) must be the same as those used for samples.


233

3.5.2. Coinjection of Hyp 1. Obtain the electropherogram for the unknown samples by performing the separation as described in Subheading 3.4. 2. Add 10 µL 2.0 mM Hyp to the sample and run the separation again. 3. Compare the two electropherograms. The peak with substantially increased area and height corresponds to the Hyp peak.

3.5.3. Standard Curve 1. Prepare a set of at least 6 Hyp dilutions covering the range of 0.05 to 2.0 mM. 2. Add 10 µL 2.0 mM thymine to 30 µL of each standard concentration and mix. 3. Perform the separations as described in Subheading 3.4. using a constant injection volume (e.g., 10 nL). 4. Determine the Hyp peak area ratio for each standard concentration by dividing the peak area for Hyp by that of the thymine peak. 5. Plot the Hyp peak area ratio vs the Hyp concentration for each point. Use a linear regression analysis of the data points. The slope of this line (Hyp concentration/peak area ratio) represents the Conversion Factor. The concentration of Hyp in the unknown sample is calculated from the equation below. Hyp peak area ratio in unk × Conversion Factor (Hyp conc/peak area) = Hyp Conc 6. To determine the amount (moles) of Hyp in the unknown sample, the total volume of the hydrolyzed aliquot must be used. For example, assuming that an aliquot corresponding to 1.0 mg of tissue was hydrolyzed, labeled, and finally dissolved in a final volume of 100 µL then the total amount of Hyp in that sample is calculated as follows: Hyp Conc × injection volume (nL) × dilution factor (100 µL/10 nL) = moles Hyp Because the amount of tissue hydrolyzed is known, the Hyp content can also be expressed as moles of Hyp per mg tissue.

4. Notes 1. Amine capillary preparation: Since the time of our original investigations, the commercial availability of amine coated capillaries has significantly increased. The quality and consistency of prepared capillaries is well worth the cost. Thus, unless the investigator is very familiar with this technology, it is recommended that amine coated capillary be obtained through a commercial vendor. 2. PTH Labeling of Hyp Standards: The labeling reaction of Hyp isomers or Pro with PITC is rather easy to do. Precautions include flushing the reaction mixture with nitrogen and protecting it from light. UV light accelerates the degradation of PTH derivatives. Storage should be in a light protected vial at –20° to –70°C. The shelf life of PTH-Hyp was only 2–3 wk. Additional peaks representing breakdown products were observed in CE separations with increasing storage.

234

Chu and Zeece

3. This method describes procedures for application of CE to the determination of Hyp isomers and Pro in muscle tissue. It is also be applicable to the analysis of these amino acids in other tissues. Additional considerations that might be helpful in performing Hyp analysis are addressed below. 4. Whole muscle samples or isolated collagen fractions can be analyzed by this procedure. The protocol has a number of steps, but is faster and requires less sample than other (HPLC) methods. The sensitivity of this method was found to be sufficient such that determinations could be made in as little as 1 mg of sample (9). 5. A critical step for determination of Hyp in muscle samples of involves labeling of amino acids in hydrolyzed sample (Subheading 3.3.). This a two-step procedure in which amino acids (all except Hyp and Pro) are reacted with OPA. The OPA amino acids are then precipitated with 1 M 1.0 M NaH2PO4, pH 2.5. The insoluble amino acid derivatives are then removed by centrifugation the supernatant containing imino acids (Hyp and Pro) is subsequently reacted in the second step with PITC. However, precipitation of OPA derivatized amino acids may not be complete with NaH2PO4 in some samples. Alternatively, we have found that substitution of OPA with Naphthalenedicarboxaldehyde results in more efficient precipitation with NaH2PO4 and greater solubility for Hyp and Pro. 6. The CE separation and determination of Hyp and Pro in biological samples is relatively straightforward and faster than many HPLC approaches because the smaller sample requirements result in faster processing time. For example, this procedure contains a number of steps in which drying of the sample is necessary. Because the CE procedure requires only 100–150 µL hydrolysis aliquots, the time required to process the sample in substantially reduced. We have also successfully performed Hyp analysis with 1–5 mg samples using microwave hydrolysis containing 200–500 µL of HCl, to further reduce the processing time. 7. A critical factor in the quantitation of Hyp in biological samples by CE is the requirement for an internal standard. CE uses very small sample volumes (nL) and thus slight differences in the injection volume can result in large variations for Hyp peak areas. The use of internal standard peak area ratio method eliminates the errors caused by injection differences. The thymine peak also provides a reference point (migration time) for assessing the consistency of electropherograms.

References 1. Grossman, P. D. (1992) Factors affecting the performance of capillary electrophoresis separations: joule heating, electrosmoosis, and zone dispersion, in Capillary Electrophoresis: Theory and Practice (Grossman, P. J. and Colburn, J. C., eds.), Academic, London and New York, pp. 3–44. 2. Kuhn, R. and Hoffstetter-Kuhn, S. (1993) Capillary Electrophoresis: Principles and Practice. Springer Laboratory, New York. 3. Karger, B. L. and Foret, F. (1993) Capillary electrophoresis: introduction and assessment, in Capillary Electrophoresis Technology (Guzman, N., ed.), Marcel Dekker, New York, pp. 3–64.


235

4. Khur, W. (1993) Separation of small organic molecules by capillary electrophoresis, in Capillary Electrophoresis: Theory and Practice (Camilleri, P., ed.), CRC, Boca Raton, FL, pp. 65–116. 5. Morning, S. E. (1996) Buffers, electrolytes and additives for capillary electrophoresis, in Capillary Electrophoresis in Analytical Biotechnology (Righetti, P., ed.), CRC, Boca Raton, FL, pp. 37–60. 6. Otsuka, K., Terabe, and S., Ando, T. (1985) Electrokinetic chromatography with micellar solutions, separation of phenylthiohydantoin-amino acids. J. Chromatog. 332, 219–226. 7. Matsubara, N. and Terabe, S. (1996) Micellar electrokinetic chromatography in the analysis of amino acids and peptides, in Capillary Electrophoresis in Analytical Biotechnology (Righetti, P., ed.), CRC, Boca Raton, FL, pp. 155–182. 8. Kuhn, R. and Hoffstetter-Kuhn. (1993) Capillary Coatings in Capillary Electrophoresis: Principles and Practice. Springer Laboratory, New York, pp. 162–180.. 9. Chu, Q., Evans, B. T., and Zeece, M. G. (1996) Quantitative separation of 4-hydroxyproline from skeletal muscle collagen by micellar electrokinetic capillary electrophoresis. J. Chromatog. B 692, 293–301. 10. Light, N. and Champion, A. E. (1984) Characterization of muscle epimysium, perimysium, and endomysium collagens. Biochem. J. 219, 1017–1026. 11. Yaegaki, K., Tonzetich, J., and Ng, A. S. K. (1996) Improved high performance liquid chromatography method for quantitation of proline and hydroxyproline in biological materials. J. Chromatog. 356, 163–170.

Total Plasma Homocysteine HPLC Analysis

237

17 Total Plasma Homocysteine Analysis by HPLC with SBD-F Precolumn Derivatization Isabella Fermo and Rita Paroni 1. Introduction Over the last two decades, various studies have shown that moderate and persistent hyperhomocysteinemia is implicated in the development of atherosclerosis, which is responsible for 50% of all mortality and morbidity in Western countries. Considering that the most traditional risk factors for heart disease and stroke, such as plasma lipids, cigarette smoking, hypertension, obesity, and diabetes, only account for 50% of cardiovascular disease (1,2), one can understand the reason why homocysteine (Hcy) measurement is included in the list of tests for investigating the causes of atherosclerosis and thrombosis. One of the major problems encountered in studies on the potential atherogenic role of Hcy was the development of an accurate and simple assay, capable of screening, in a normal population, subjects having a congenital predisposition to occlusive vascular disease. Several approaches have been described in literature for measuring total plasma homocysteine (tHcy), which is defined as the sum of free and protein-bound homocysteine, homocystine, and homocysteine-cysteine mixed disulfide. These procedures involve, after a reduction step, the use of gas-chromatography-mass spectrometry (GC-MS) (3), radioenzymic assay (4), and high-performance liquid chromatography (HPLC). The latter, which is the most widely applied, may be coupled with (1) spectrophotometric detection in the visible range (amino acid analyzer) after postcolumn ninhydrin derivatization (5); (2) direct electrochemical detection (6); or (3) spectrofluorimetric detection after precolumn derivatization with different labeling agents such as monobromobimane (mBrB) (7), o-phthaldialdehyde (OPA) (8), or halogensulfonyl benzofurazans (ABD-F and SBD-F) (9,10). From: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

237

238

Fermo and Paroni

Among the HPLC procedures for tHcy analysis in plasma, methods based on SBD-F (ammonium-7-fluorobenzo-2-oxa-1,3-diazole-4-sulphonate) reaction appear to be attractive for routine use in clinical chemistry laboratories. SBD-Hcy adduct is very stable (1 wk at 4°C) (see Note 1) (11) and SBD-F being a thiolspecific labeling molecule, the derivatization yields a very clean chromatogric profile in comparison to mBrB, which produces interfering fluorescence byproducts (7), or to OPA, which reacts unspecifically with all the amino groups of the sample (8). We have described here a procedure that involves reduction of disulfide bonds by sodium borohydride (NaBH4), use of mercaptopropionylglycine as internal standard (12), precolumn derivatization with SBD-F, and isocratic analysis by reversed-phase (RP) HPLC with fluorescence detection. 2. Materials

2.1. Equipment 1. Polypropylene micro tubes (2 mL) with screw-cap (Sarstedt, Numbrecht, Germany). 2. Microfuge model 11 (Beckman, Palo Alto, CA). 3. Block heater (Asal, Milano, Italy) or oven. 4. HPLC monopump model 116 with a solvent selector valve (Beckman, Palo Alto, CA) (see Note 2). 5. Fluorimetric detector model RF 551 (Shimadzu, Kyoto, Japan). 6. Autosampler model 507 (optional device) (Beckman). 7. Column: Ultrasphere ODS (150 × 4.6-mm id, 5 µm) (Beckman). 8. Guard-column: Lichrospher 100 RP–18 (40 × 4-mm id, 5 µm) (Merck, Darmstadt, Germany).

2.2. Reagents 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

NaBH4. Dimethyl sulfoxide (DMSO). 7-fluorobenzo-2-oxa-1,3-diazole-4-sulphonate (SBD-F). Boric acid. ethylenediaminetetracetic acid (EDTA). L-homocystine (Hcy-Hcy). Sodium acetate. Sodium hydroxide (NaOH). 70% perchloric acid (PCA). Methanol (CH3OH). Acetonitrile (CH3CN). Glacial acetic acid. Mercaptopropionylglycine (IS). Bidistilled water (H2O).


239

2.3. Solutions 1. Reducing solution: Prepare 3 mol/L NaBH4 solution by dissolving 0.285 mg of NaBH4 in 2.5 mL of NaOH 0.1 mol/L, then mix with DMSO (2:1, v/v) and store in a glass vial at 4°C. Prepare this solution fresh every week. 2. Deproteinizing solution: Prepare 0.6 mol/L PCA solution by mixing 25 mL PCA (70%) with 0.186 g EDTA and diluting to 500 mL with H2O. This solution is indefinitely stable when stored at 4°C (see Note 3). 3. Buffer solution: Prepare 0.2 mol/L boric acid solution with 4 mmol/L EDTA, pH 9.5 by dissolving 12.37 g boric acid and 1.49 g EDTA in 1 L H2O. Warm to dissolve. Adjust to pH 9.5 with concentrated NaOH while still warm. Cool to room temperature before use and store this solution at room temperature. 4. Neutralizing solution: Prepare NaOH 2 mol/L by dissolving 80 g NaOH in 1 L H2O. 5. Derivatizing solution: Prepare 4.2 mmol/L SBD-F solution by mixing 1 mg SBDF with 1 mL of buffer solution. Store at 4°C in a dark vial for 1 d. 6. Internal standard solution: Mix 1 mg of IS with 1 mL HCl 0.1 mol/L and 9 mL H2O. Store 200-µL aliquots at –20°C up to 6 mo. To use, dissolve each aliquot with 600 µL solution 3 diluted 1:1 with H2O. Store at 4°C for 1 wk. 7. Homocystine standard solution: Prepare solution 200 µmol/L by dissolving in a 100 mL volumetric flask 5.4 mg of Hcy-Hcy in 5 mL HCl 0.1 mol/L and diluting to 100 mL with H2O. Freeze 300-µL aliquots at –20°C for 6 mo. Use each aliquot only once and then discard. 8. Stock HPLC eluents: Prepare sodium acetate solution 0.2 mol/L by dissolving 27.22 g sodium acetate trihydrate in 1 L H2O. Dilute 11.5 mL glacial acetic acid with 1 L H2O and adjust to pH 4.0 with the sodium acetate solution (≈3:1) to obtain 0.2 M sodium acetate buffer. Store all these solutions at 4°C. 9. Working HPLC eluents: Mobile phase A: add 20 mL of CH3OH to 980 mL of buffer 8 (2% CH3OH). Mobile phase B: add 300 mL of CH3OH to 700 mL of solution 8 (30% CH3OH).

3. Methods

3.1. Preanalytical Steps Collect whole blood specimens in sterile Vacutainer® tubes containing sodium citrate (Becton Dickinson, Rutherford, NJ). Keep the samples on ice until centrifugation at 3000g for 10 min at 4°C, then immediately freeze the plasma specimens at –20°C until analysis (see Note 4).

3.2. Procedure (see Flow Chart 1) 1. Reduction step. Mix 100 µL plasma sample with 50 µL of internal standard solution and 20 µL of NaBH4 solution (reducing solution) in a polypropylene tube with screw-cap. Incubate at 50°C for 30 min (see Note 5).

240

Fermo and Paroni

Flow Chart 1 SBD-F Precolumn Derivatization Method • Plasma collection

⇒

• IS addition • Reduction of disulfide bonds

⇒ ⇒

• Plasma deproteinization

⇒

• Plasma neutralization • Derivatization • HPLC separation

⇒ ⇒ ⇒

• Fluorimetric detection

⇒

Anticoagulant: sodium citrate, eparin, or EDTA - Blood sample in ice - Prompt separation of plasma 100 µL mercaptopropionylglycine solution by 20 µL 3 mol/L NaBH4 solution (30 min at 50°C) by 100 µL 0.6 mol/L PCA in 1 mmol/L EDTA by 20 µL 2 mol/L NaOH by 50 µL SBD-F solution (60 min at 60°C) isocratic elution (14 min); reversed-phase C18 column λex = 385 nm; λem = 515 nm

2. Deproteinization step. Add 100 µL 0.6 mol/L PCA (solution 2). Vortex, then centrifuge at 12000g for 10 min. 3. Derivatization step. Transfer 100 µL of supernatant in an Eppendorf tube and add 200 µL of 0.2 mol/L borate buffer pH 9.5 (buffer solution), 20 µL of 2 mol/L NaOH (neutralizing solution), and 50 µL of SBD-F solution (derivatizing solution). Vortex and then incubate the samples at 60°C for 1 h. Remove and allow vials to cool to room temperature before injecting (see Note 6). 4. Transfere derivatized samples into the autosampler vials (optional device) and inject 10 µL into the HPLC (see Notes 7 and 8).

3.3. Calibration Curve 1. For each batch of samples to be run on HPLC, dilute the 200 µmol/L homocystine stock solution to get the final concentrations: 100, 50, and 25 µmol/L (use solution 3 diluted 1:1 with H2O). 2. Spike 100 µL of pooled plasma with 10 µL of each of these hcy-hcy standard solutions to obtain a calibration curve in the 5, 10, 20, 40 µmol/L range (Hcy molar equivalents) (see Notes 7 and 9). Plot the relative fluorescence intensities of SBF-Hcy/IS (y) as a function of Hcy concentrations (x), and perform the least square analysis to obtain the regression equation. 3. The unknown tHcy concentration in a plasma sample is determined by this calculation: [tHcy]sample µmol/L = (Hcy peak area : IS peak Area)sample/slope of calibration curve


241

Fig. 1. Chromatograms of: (—) plasma sample with tHcy level of 23.0 µmol/L; (---) Bio-Rad Liquicheck™ Homocysteine Control-2 (lot. no. 18312, Bio-Rad Diagnostic Group, Irvine, CA). The separation was carried out on a Beckman Ultrasphere ODS column (150 × 4.6-mm id; 5 µm). (peak 1 = cysteine; peak 2 = homocysteine; peak 3 = cysteinylglycine; peak 4 = glutathione; peak 5 = mercaptopropionylglycine).

3.4. Chromatographic Analysis 1. Fluorimetric detector: λex = 385 nm; λem = 515 nm. 2. Run time: 14 min. 3. Equilibrate the column with 100% mobile phase A at 1.2 mL/min and 0.2 min after the analysis started, switch the selector valve to 100% eluent B. 4. At 4 min increase flow rate to 1.4 mL/min over 0.5 min, and at 4.5 min, switch back the valve to 100% eluent A. 5. At 12 min, decrease the flow rate to 1.2 mL/min in 0.5 min and at 14 min you are ready for the next injection.

242

Fermo and Paroni HPLC Run Program Time Flow (min) (mL/min) 0 1.2 0.2 1.2 4.0 1.4 4.5 1.4 12 1.2 14 STOP

Duration (min) — — 0.5 — 0.5 —

Mobile phase A B B A A A

The chromatogram in Fig. 1 shows the following peaks: cysteine (1), homocysteine (2), cysteinylglycine (3), glutathione (4), and mercaptopropionylglycine (5).

4. Notes 1. The fluorescence SBD-Hcy adduct, being photosensitive, has to be stored in the dark where is stable for 1 wk at 4°C (11). 2. This method requires the use of a simple monopump HPLC, but with a solvent selector valve. If the solvent switching is not available, the IS will be eluted later and the analysis will last too long. As an alternative, the method can be performed without the addition of IS, and the chromatographic separation will be performed using only the mobile phase A at flow rate 1.2 mL/min. In this case, we have to be careful, above all, during the reduction step and the use of a surface active agent is recommended. In addition, we suggest running the plasma calibration curve at least three times per day in order to check the performance of own detector system. 3. If the deproteinizing solution shows an amount of precipitate on the bottom of the glass vial, we are probably using out-of-date reagents. The reductant must to be completely dissolved in the vial before its use. 4. We adopted sodium citrate as anticoagulant as it is commonly used by the Coagulation Service of our Institute; no differences in tHcy plasma levels were found after EDTA or heparin treatment (8). It is important to know that differences in blood-sample handling can cause significantly different results in tHcy measurement. During whole blood storage at room temperature, Hcy concentration tends to increase over time because of the protracted production and release of Hcy by blood cells. This process is slowed down when blood samples are left on ice. There are no problems for plasma samples because tHcy in this biological fluid is stable for at least 4 d at room temperature, for several weeks at 4°C and for years at –20°C (7). 5. Sodium borohydride is a potent reductant, but because this reaction involves the formation of gas and sample foaming, the use of surface active agents such as amyl alcohol is also suggested. In our method, this addition is avoided thanks to (1) the use of 2-mL conical bottom tubes with screw-caps and (2) the addition of an internal standard. Pay attention during PCA deproteinization also: as the addition of an acid sparks off sample foaming. For this reason, we prefer to utilize tubes with an internal volume of 2 mL, at least.


243

6. Derivatization reaction can be directly performed in the autosampler vials if these tubes are heat-resistant and suitable for the block heater (incubation equipment). 7. The lowest limit for reliable quantification (LOQ) of the method is 0.85 µmol/L, corresponding to 1.7 pmol injected (12), and the analysis is carried out by utilizing a simple fluorometric detector on line with the HPLC apparatus. 8. The normal ranges for plasma tHcy differ from one laboratory to another, and these discrepancies may be related to the procedure or technique used or the criteria chosen to select the control subjects. However, tHcy plasma levels between 5.0 and 15.0 µmol/L are usually considered as normal (13). By applying this procedure, we found the following reference values: 7.65 ± 2.36 and 8.9 ± 2.14 µmol/L for females (n = 40) and males (n = 40), respectively (8). 9. Plasma calibration curve must be used instead of aqueous calibrators because the latter give a linear equation with a different slope (–20%), by showing highlighting the notable influence of the matrix on SBD-F thiol derivatization (9,12). We can use Hcy standard instead of Hcy-Hcy, but, in this case, care must be taken with the purity of the Hcy powder (14). All standards must be stored at –20°C.

References 1. Masser, P. A., Taylor, L. M. D., and Porter, J. M. (1994) Importance of elevated plasma homocysteine levels as a risk factor for atherosclerosis. Ann. Thorac. Surg. 58, 1240–1246. 2. Stampfer, M. J. and Malinow, M. R. (1995) Can lowering homocysteine levels reduce cardiovascolar risk? N. Engl. J. Med. 332, 328–329. 3. Sass, J. O. and Endres, W. (1997) Quantitation of total homocysteine in human plasma by derivatization to its N(O, S)-propoxycarbonyl propyl ester and chromatography-mass spectrometry analysis. J. Cromatog. A 776, 342–347. 4. Frantzen, F., Faaren, A. L., Alfheim, I., and Nordhei, A. K. (1998) Enzyme conversion immunoassay for determining total homocysteine in plasma or serum. Clin. Chem. 44, 311–316. 5. Andersson, A., Brattstrom, L., Isaksson, B., and Israelsson, B. (1989) Determination of homocysteine in plasma by ion-exchenge chromatography. Scand. J. Clin. Invest. 49, 445–449. 6. Malinow, M. R., Kang, S. S., and Taylor, L. M. (1989) Prevalence of hyperhomocyst(e)inemia in patients with peripheral arterial occlusive disease. Circ. Res. 79, 1180–1188. 7. Fiskerstrand, T., Refsum, H., Kvalheim, G., and Ueland, P. M. (1993) Homocysteine and other thiols in plasma and urine:automated determination and sample stability. Clin. Chem. 39, 263–271. 8. Fermo, I., Arcelloni, C., De Vecchi, E., Vigano’, S., and Paroni, R. (1992) HPLC method with fluorescence detection for the determination of total homocyst(e)ine in plasma. J. Chromatog. (Symp) 593, 171–176. 9. Vester, B. and Rasmussen, K. (1991) HPLC method for rapid and accurate determination of homocysteine in plasma and serum. Eur. J. Clin. Chem. Clin. Biochem. 29, 549–554.

244

Fermo and Paroni

10. Feussner, A., Rolinski, B., Weiss, N., Deufel, T., and Rosher, A. (1997) Determination of total homocysteine in human plasma by isocratic HPLC. Eur. J. Clin. Chem. Clin. Biochem. 35, 687–691. 11. Imai, K., Toyo’oka, T., and Watanabe, Y. (1983) A novel fluorogenic reagent for thiols:ammonium 7-fluorobenzo-2-oxa-1, 3-diazole-4-sulfonate. Anal. Biochem. 128, 471–473. 12. Fermo, I., Arcelloni, C, Mazzola, G., D’Angelo, A., and Paroni, R. (1998) HPLC method for measuring total plasma homocysteine levels. J. Chromatog. B 719, 31– 36. 13. Ueland, P. M., Refsum, H., Stabler, S. P., Malinow, M. R., Andersson, A., and Allen, R. H. (1993) Total homocysteine in plasma or serum: methods and clinical application. Clin. Chem. 39, 1764–1779. 14. Dudman, N. P. B., Guo, X. W., Crooks, R., Xie, L., and Silberberg, J. S. (1996) Assay of plasma: light sensitivity of the fluorescence SBD-F derivative, and use of appropriate calibrators. Clin. Chem. 42, 2028–2032.

Assaying Early and Crosslinked Glycation Products

245

18 Determination of Early Glycation Products by Mass Spectrometry and Quantification of Glycation Mediated Protein Crosslinks by the Incorporation of [14C]lysine into Proteins Malladi Prabhakaram, Beryl J. Ortwerth, and Jean B. Smith 1. Introduction The Maillard reaction, popularly known as nonenzymatic glycosylation (NEG) or glycation, is a complex chemical reaction and occurs in vivo between reactive aldose or ketose sugars and protein-bound free amino groups (1). NEG has been implicated in diabetic or age-related complications (2), Alzheimer’s disease (3), and also in cataract formation (4). In vivo, any protein with free amino groups can react with reducing sugars via the Maillard reaction. However, the extent of damage caused by NEG is amplified in diabetic tissues (because of the elevated levels of blood sugar levels) and also in proteins with a long half-life like collagen and lens crystallin proteins (1,2). Therefore, determination of early glycation products and protein crosslinks produced because of glycation will be a significant aspect to investigate the extent of damage caused by NEG in vivo. As shown in Fig. 1, the Maillard reaction involves the initial interaction of a reducing sugar (like glucose in Fig. 1) with α-/ε- amino groups of proteinbound amino acids, with the ε- amino group on lysine as the preferential site. An initial labile Schiff base formed during the reaction undergoes Amadori rearrangement and yields a stable ketoamine derivative. Following dehydration and subsequent chemical modification, some of the Amadori products accumulate with time as advanced glycation end products (AGEs). Though it has been proposed that AGEs have a significant impact in age-related or diabetic

From: Methods in Molecular Biology, vol. 159: Amino Acid Analysis Protocols Edited by: C. Cooper, N. Packer, and K. Williams © Humana Press Inc., Totowa, NJ

245

246

Prabhakaram et al.

Fig. 1. Reaction scheme for the Maillard reaction (nonenzymatic glycation). Glucose is shown as the model sugar and lysine is shown as the model amino acid.

complications (1–3), it is still a debate to understand the mechanism involved in the development of glucose dependent diabetic complications. This is partly because of the slow reaction property of glucose with model proteins in vitro (5). Previously, we have shown that the oxidation products of ascorbic acid (which is relatively higher than glucose in human lens) can glycate and crosslink lens proteins at an accelerated rate compared to glucose under similar in vitro conditions (6–8), suggesting a possible role for ascorbate mediated glycation in vivo. Characterization and quantification of glycation crosslinks has been the subject of study for decades. Several methods have been developed in the past to determine early and advanced glycation products. These methods include borohydride reduction (9), thiobarbituric acid assay (10), the fructosamine assay (11), periodate assay (12), size-exclusion chromatography (13), ion–exchange chromatography (14), boronate-affinity chromatography (15), and immuno detection (16). With the exception of immunological methods, almost all the glycation assay methods are based on chemical methods, and are nonspecific. Additionally, in vitro assay methods do not guarantee the detection of similar compounds in vivo. For example, some of the reagents used in borohydride reduction, thiobarbituric acid assay, acid/base hydrolysis, or periodination can nonspecifically interact with nonglycated products in the assay system and generate faulty reaction products that are counted as glycated products. Other examples are the fructosamine assay and phenylboronate assay, which are based on the detection of serum albumin reacted with glucose (aldose sugar). Both these assay methods failed to detect serum albumin glycated by fructose (ketose sugar) in diabetics (17). Although measurement of fluorescent compounds like pentosidine (18) in glycated proteins is an alternate choice, not all AGEs (e.g., carboxymethyl lysine, ref. 19) are fluorescent in nature. Another drawback in estimation of fluorescent AGEs is the highly crosslinked or insol-


247

uble nature of glycated proteins in diabetic or aged tissues. Therefore, alternate assay methods to detect and quantitate the early and advanced glycation end products represent a needed and valid approach to estimate the damage caused by nonenzymatic glycation in vivo. Recently, we have shown that mass spectrometry (MS) can be effectively used to detect and determine the early glycation products formed in a model glycation system (20). MS has successfully been used to precisely determine glycated proteins or peptides in model glycation systems (21–23) (see Note 1). Although sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) is a widely used method to observe crosslinked proteins, high molecular-weight proteins (>100 kDa) that are formed because of glycation are poorly resolved on polyacrylamide gels. These high molecular-weight proteins compress as a single band on the top of the gel and scan as a single peak by densitometric gel scanners. In order to quantitate glycation crosslinks, we have developed an in vitro assay method based on the incorporation of [14C]lysine into proteins during glycation (24). Because AGE inhibitors like aminoguanidine and semicarbazide inhibit the incorporation of [14C]lysine into proteins (24), this assay can be used to compare the crosslinking ability of various sugars, and it can also be used to test the ability of specific amino acids to act as donors for crosslinking to protein (see Note 2). In this chapter, we describe a method to determine early glycation products by MS and also a method to quantitate the glycation crosslinks in model proteins by the incorporation of [14C]lysine into proteins. 2. Materials

2.1. Equipment 1. High-performance liquid chromatography (HPLC) system + ultraviolet (UV) detector. 2. Vydac C18 HPLC column (10 × 250 mm) (Vydac, The Separations Group, Hesperia, CA). 3. Nylon filters (Gelman Scientific, Ann Arbor, MI). 4. 1.5-mL sterile amber vials (Pierce Chemicals, Rockford, IL). 5. Metal ion-free Eppendorf tubes (Perfector Scientific, Atascadero, CA). 6. FAB-MS (Varian, Netherlands). 7. ESI-MS (Varian).

2.2. Reagents 1. Diethylenetriaminepentaacetic acid (DTPA), [1- 12C]L-threose and other biochemicals can be obtained from Sigma Chemical Co. (St. Louis, MO). [1-13C]Lthreose can be obtained from Omicron Chemicals, South Bend, IN, and α-N-tBoc-lys-ala-ala peptide can be obtained from Bachem, Inc. (Philadelphia, PA).

248

Prabhakaram et al.

2. Radioactive amino acids like [14C]lysine (specific activity 165 mCi/mmol) and [ 14C]leucine (specific activity 315 mCi/mmol) can be obtained from NEN Research Products (Boston, MA) or ICN Biomedicals Inc. (Costa Mesa, CA). Ultrapure lysozyme is from AMRESCO, Solon, OH. Model sugars like glucose, ascorbic acid, dehydroascorbic acid, or L-threose and RNase can be obtained from Sigma Chemical Co. Trichloroacetic acid (TCA) can be obtained from Fisher Scientific (Pittsburgh, PA). 3. Metal-free deionized water should be used to prepare buffers and other reagents.

2.3. Solutions 1. Phosphate-DTPA buffer: 0.1 M sodium phosphate buffer, pH 7.0, containing 0.1 mM DTPA. 2. Stock A: 1 M Boc-KAA peptide in phosphate -DTPA buffer. 3. Stock B: Threose in phosphate -DTPA buffer. 4. Solvent A: 0.1% TFA.

3. Methods

3.1. Glycation of t-Boc-lys-ala-ala (Boc-KAA) Peptide by Threose [12C and 1–13C] 1. Prepare 0.1 M sodium phosphate buffer, pH 7.0 containing 0.1 mM DTPA as the chelator (referred as phosphate-DTPA buffer) in deionized water. Prepare stock solutions of 1 M Boc-KAA peptide and threose in phosphate -DTPA buffer just before use. 2. Prepare a reaction mixture containing peptide and threose (1:10 on a molar basis) by mixing 20 µL of 1 M peptide and 200 µL of 1 M threose in 1 mL of phosphateDTPA buffer. Also prepare control reaction mixtures without peptide and without sugar in 1 mL of phosphate-DTPA buffer. 3. Immediately after the addition of reactants, filter the reaction mixtures through 0.2-µm sterile nylon filters (Gelman Scientific) into sterile amber vials of 1.5-mL capacity (Pierce Chemicals). Use a laminar flow hood for this purpose. 4. Incubate the vials with reaction mixtures in an incubator maintained at 37°C. 5. To observe the reaction products that accumulate with time, withdraw aliquots (equivalent to approx 100 µg peptide) at different intervals (0–72 h) (see Note 3) and freeze the aliquots immediately at –80°C.

3.2. Preparation of Samples for Mass Spectrometry 1. Thaw aliquots frozen at –80°C (step 5, Subheading 3.1.) at room temperature and inject on a semipreparative Vydac C18 RPHPLC column (10 × 250 mm), which has been preequilibrated with 0.1%TFA in water (solvent A) (see Note 4). Set the absorbance on UV detector at 210 nm to monitor the absorbance of the peptide and reaction products. 2. Elute reaction products that are not bound to HPLC column in 5 column volumes of solvent A. This is followed by the elution of reaction products that are bound


249

to HPLC column in approx 10 column volumes of solvent B (60% acetonitrile and 0.1% TFA in water). 3. For accurate mass determination, collect the reaction products in metal-free (see Note 5) Eppendorf tubes (1.5-mL/tube/min) and combine the fractions with maximum absorbance at 210 nm (see Note 6).

3.3. FABMS Analysis of Modified Peptides 1. Dissolve the dried samples (step 3, Subheading 3.2.) in 5–10 µL of methanolwater and add 90% formic acid (Spectrapure) to a final concentration of 22% (see Note 7). Remove a 1.0-µL aliquot and mix immediately with a drop of a 1-thioglycerol and place on the MS instrument probe (see Note 8). 2. Insert the probe into a mass spectrometer and record spectra over a m/z range of 300–2000 at a resolution of 2000 (see Notes 9 and 10).

3.4. Electrospray Ionization Mass Spectrometry (ESIMS) of BocKAA Peptide Modified During Glycation 1. Dissolve the samples obtained from HPLC (step 3, Subheading 3.2.) in a solution of 50:50 acetonitrile:water with 0.1% TFA (see Note 7) and deliver to the analyzer at a flow rate of 5 µL/min in the same solution (see Note 11). 2. Record the spectra over a m/z range 300–2000 and analyze the data with massLinx software or any other appropriate software.

3.5. Quantification of Glycation Crosslinks in Proteins by the Incorporation of [14C]Lysine into Proteins During Glycation We describe below an assay method to determine the glycation-dependent crosslinks formed between [14C]lysine and lysozyme in the presence of a model sugar, L-threose (see Note 12). To measure the reactivity of the α-amino group alone with lysozyme/RNase during glycation, [14C]leucine can be used instead of [14C]lysine in the reaction mixture. Adequate care should be taken to handle and dispose of any radioactive waste according to NRC guidelines. 1. Prepare 0.1 M sodium phosphate buffer, pH 7.0, containing 0.1 mM DTPA as the chelator (referred as phosphate-DTPA buffer) in deionized water. Just before use, prepare stock solutions of lysozyme or RNase (25 mg/mL) and L-threose or reactive sugar of choice (1 M) in phosphate-DTPA buffer. 2. Dilute the stock solution of [14C]lysine to a working stock of 50 µCi/100 µL in deionized water and keep on ice. 3. A typical 1-mL reaction mixture is prepared as described below (see Note 13). Lysozyme/RNase (25 mg/mL) 200 µL

[14C]lysine 50 µCi/100 µL 10 µL

Phosphate-DTPA buffer 770 µL

L-threose (1 M) 20 µL

4. Immediately after the addition of L-threose, sterile filter each reaction mixture through a 0.2-µm nylon filter into an amber vial (1.5-mL capacity) and incubate in an incubator maintained at 37°C.

250

Prabhakaram et al.

5. Withdraw aliquots in duplicate at different intervals from 0–72 h and immediately freeze the aliquots at –80°C until use.

3.6. Measurement of Radioactivity in the Aliquots Withdrawn from Reaction Mixtures Because free lysine readily adheres to proteins and methods like dialysis or size-exclusion chromatography do not completely remove lysine nonspecifically bound to proteins, we routinely used the filter disk method of Mans and Novelli (26) to determine the protein-bound radioactivity. According to this method, proteins will be precipitated by cold TCA on a 3 MM Whatman filter paper disk and any unbound or low molecular-weight compounds will be washed off prior to determine the protein-bound radioactivity. The proteins bind to the fiber of the filter disk and remain on the filter paper during the washing procedure with TCA. 1. Aliquots frozen at –80°C should be thawed prior to use. Spot 10 mL of each aliquot (in triplicate) on 3 MM Whatman filter paper disks and air dry the filter paper disks under a fume hood (filter paper disks can conveniently suspended with pins on a Styrofoam box lid to air dry). 2. After the reagents have been absorbed into the disks, drop the filter paper disks with pins still inserted into a solution of ice-cold 10% TCA while stirring with a magnetic stirrer and wait 20 min (see Note 14). A perforated 250-mL Plastic beaker (a cork borer can be used to make holes in the plastic beaker) can be used to hold the filter paper disks and place the plastic beaker in a 500-mL glass beaker containing 10% TCA. 3. Transfer the plastic beaker with filter paper disks to a glass beaker containing 5% TCA (room temperature) and stir for 20 min. Following this, transfer the plastic beaker with filter paper disks to a glass beaker containing 5% TCA maintained at about 75°C and stir for 15 min. Hot TCA assures the measurement of incorporation into protein only. 4. Transfer the plastic beaker with filter paper disks to a mixture of ethanol-anhydrous ether (2:1) under stirring and wait for 15 min. This step is required to remove residual TCA from the filter paper disks. 5. Transfer the filter paper disks carefully to a plastic weighing boat or to a sheet of aluminium foil and air-dry the filter paper disks in the fume hood or under a heat lamp. 6. Remove the pins, transfer the dried filter paper disks to a vial containing radioactive scintillation cocktail (e.g., Ultima gold, Packard, Meriden, CT) and count radioactivity in a liquid scintillation counter. Calculate the amount of radioactivity present on each disk and determine the amount incorporated into lysozyme or RNase at different intervals during glycation reaction.

4. Notes 1. Determination of molecular weights of glycated peptides by either fast atom bombardment mass spectrometry (FABMS) or by electrospray ionization mass spec-


2.

3.

4.

5.

6.

7. 8.

9.

10.

251

trometry (ESIMS) is an effective method of establishing whether a peptide has become glycated, how many sugar moieties are attached to the peptide, and whether there is evidence of crosslinks. Typical mass accuracy for determinations by both FABMS and ESIMS is about ± 0.2 Daltons. Although both techniques can be used, each has its advantages and disadvantages. FABMS is customarily used for peptides with molecular masses of 250–4000. The mass range for ESIMS is much larger. ESIMS can be used to determine molecular masses of both peptides and proteins. Although [14C]lysine can be used to measure glycation dependent crosslinking of [14C]lysine to model proteins, N-α-acetyl[14C]lysine can be used to estimate the glycation crosslinks formed because of the specific interaction of ε-amino group of lysine with proteins during glycation (27). To measure specific lysine-amino acid crosslinks, homopolymers of amino acids like polylysine, polyarginine or polyhistidine can also be used. L-threose is a highly reactive tetrose sugar and its reaction with free amino groups is generally completed in only a few days (25). Therefore, L-threose is selected as a model sugar in the experiments. If threose is substituted with sugars like glucose or fructose, aliquots should be withdrawn at weekly intervals or until the reaction products reach to a plateau. FABMS or ESIMS require that the sample should be free of nonvolatile salts containing cations such as sodium and potassium. These cations can replace the protons on some peptides, leading to many additional charged states. For this reason, it is recommended that the last step in the isolation of the peptides be RP-HPLC. Dialysis usually does not adequately remove interfering salts. Because of the extreme sensitivity to sodium ions, samples should be collected and stored in plastic tubes after the final RP desalting. Because sodium ion impurities in samples injected on MS can lead to inaccurate mass spectra, care should be taken to select metal ion-free Eppendorf tubes. We routinely used plastic tubes obtained from Perfector Scientific, Atascadero, CA. The most important information for someone submitting samples for mass spectrometric analysis is how much sample is required for analysis, and that the samples should not contain too many components and be free of salts. For both FABMS and ESIMS, the sample is usually dissolved in an acidic solvent, such as 0.1% TFA or acetic acid before introduction into the mass spectrometer. FABMS analysis is performed by mixing the dissolved sample with a matrix such as glycerol/thioglycerol or nitrobenzyl alcohol, and placing this mixture on the probe that is inserted into the instrument. The peaks caused by the matrix contribute to higher background signals, making this technique less sensitive than ESIMS, particularly at masses below 300 Daltons. We have used a range between 300–2000, because the peptide we used in this study has a molecular mass of 389. An appropriate range should be selected based on the size and mass of protein or peptide in use. One advantage of FABMS is that, in the positive ion mode, a monoprotonated peptide is obtained, so the spectrum shows an MH+ ion, which is the molecular

252

11.

12. 13.

14.

Prabhakaram et al. mass of the peptide plus one Dalton. The molecular mass of the peptide is readily apparent. In contrast, ionization of the sample in ESIMS produces several charged states of each peptide, resulting in a spectrum showing peaks with the mass/ charge ratio of each of the charged states. For ESIMS, the dissolved sample is introduced into the mass spectrometer as a continuous flow of solution. Although the amount of sample required depends very much on the sample characteristics, about one nanomole of peptide is typical for a FABMS analysis, whereas 100 picomole is usually sufficient for ESIMS. Detection limits 1000-fold lower may be achieved in special cases. One can also use radioactive-labeled proteins instead of [14C]lysine to measure the incorporation of labeled sugars into proteins. Always set up a reaction without sugar to correct for any nonspecific interaction of [14C]lysine with proteins. Also minimize the volume (50–100 µL) of samples with [14C]lysine to apply on the filter paper disks. To show that no carryover of [14C]lysine from one disk to others in between washings of the disks, include plain filter paper disks during the TCA wash procedure.

References 1. Monnier, V. M. (1989) Toward a Maillard Reaction Theory of Aging (Baynes, J. W. and Monnier, V. M., eds.). Alan R. Liss, New York. 2. Baynes, J. W. (1996) The role of oxidation in the Maillard reaction in vivo, in The Maillard Reaction: Consequences for the Chemical and Life Sciences (Ikan, R., ed.), Wiley, New York. 3. Markesbery, W. R. (1997) Oxidative stress hypothesis in Alzheimer’s disease. Free Rad. Biol. Med. 23, 134–147. 4. Monnier, V. M. and Cerami, A. (1983) Detection of nonenzymatic browning products in the human lens. Biochim. Biophys. Acta 760, 97–103. 5. Ortwerth, B. J., Speaker, J. A., Prabhakaram, M., Lopez, M., Li, E., and Feather, M. S. (1994) Ascorbic acid glycation: the reactions of L-threose in lens tissue. Exp. Eye Res. 58, 665–674. 6. Prabhakaram, M. and Ortwerth, B. J. (1991) The Glycation-associated crosslinking of lens proteins by ascorbic acid is not mediated by oxygen free radicals. Exp. Eye Res. 53, 261–268. 7. Prabhakaram, M. and Ortwerth, B. J. (1992) The glycation and cross-linking of isolated lens crystallins by ascorbic acid. Exp. Eye Res. 55, 451–459. 8. Prabhakaram, M. and Ortwerth, B. J. (1992) Glycation of MP26 and MP22 in bovine lens membranes. Biochem. Biophys. Res. Comm. 185, 496–504. 9. Kennedy, D. M., Skillen, A. W., and Self, C. H. (1993) Colorimetric assay of glycoprotein glycation free of interference from glycosylation residues. Clin. Chem. 39, 2309–2311. 10. Parker, K. M., England, J. D., Da Costa, J., Hess, R. L., and Goldstein, D. E. (1981) Improved colorimetric assay for glycosylated hemoglobin. Clin. Chem. 27, 669–672. 11. Johnson, R., Metcalf, P. A., and Baker, J. R. (1982) Fructosamine: A new approach to the estimation of serum glycosylprotein. An index of diabetic control. Clin. Chim Acta 127, 87–95.


253

12. Ahmed, N. and Furth, A. J. (1991) A microassay for protein glycation based on the periodate method. Anal. Biochem. 192, 109–111. 13. Yatscoff, R. W., Tevaarwerk, G. J. M., and McDonald, J. C. (1984) Quantification of nonenzymically glycated albumin and total serum protein by affinity chromatography. Clin. Chem. 30, 446–449. 14. Castagnola, M., Caradonna, P., Bertollini, A., Cassiano, L., Rossetti, D. V., and Salvi, M. L. (1985) Determination of the non-enzymatic glycation of hemoglobin by isoelectrofocusing of its globin chains. Clin. Biochem. 18, 327–331. 15. Middle, F. A., Bannister, A., Bellingham, A. J., and Dean, P. D. G. (1983) Separation of glycosylated haemoglobins using immobilized phenylboronic acid. Biochem. J. 209, 771–779. 16. Ikeda, K., Higashi, T., Sano, H., Jinnouchi, Y., Yoshida, M. Araki, T., Ueda, S., and Horiuchi, S. (1996) N-ε-(Carboxymethyl) lysine protein adduct is a major immunologicalepitope in protein modified with advanced glycation end products of Maillard reaction. Biochemistry 35, 8075–8083. 17. Ahmed, N. J. and Furth, A. J. (1992) Failure of common glycation assays to detect glycation by fructose. Clin. Chem. 38, 1301–1303. 18. Sell, D. R., and Monnier, V. M. (1989) Structure elucidation of a senescence crosslink from human extracellular matrix. J. Biol. Chem. 264, 21597–21602. 19. Ahmed, M. U., Thorpe, S. R., and Baynes, J. W. (1989) Identification of N-epsiloncarboxymethyllysine as a degradation product of fructoselysine in glycated protein. J. Biol. Chem. 26, 4889–4894. 20. Prabhakaram, M., Smith, J. B., and Ortwerth, B. J. (1996) Rapid assessment of early glycation products by mass spectrometry. Biochem. Mol. Biol. Int. 40, 315–325. 21. Smith, J. B., Sun, Y., Smith, D. L., and Green, B. (1992) Identification of the posttranslational modifications of bovine lens alpha B-crystallins by mass spectrometry. Protein Sci. 1, 601–608. 22. Lapolla, A., Baldo, L., Aronica, R., Gerhardinger, C., Fedele, D., Elli, G., Seraglia, R., Catinella, S., and Traldi, P. (1994) Matrix-assisted laser desorption ionization mass spectrometric studies on protein glycation 2. The reaction of ribonuclease with hexoses. Biol. Mass. Spectr. 23, 241–248. 23. Lapolla, A., Fedele, D., Seraglia, R., Catinella, S., and Traldi, P. (1994) Matrixassisted laser desorption/ionization capabilities in study of non-enzymatic protein glycation. Rapid Commn. Mass Spectr. 8, 645–52. 24. Prabhakaram, M. and Ortwerth, B. J. (1994) Determination of glycation crosslinking by the sugar-dependent incorporation of [14C]lysine into protein. Anal. Biochem. 216, 305–312. 25. Lopez, M. G. and Feather, M. S. (1992) The production of threose as a degradation product from L-ascorbic acid. J. Carbohydr. Chem. 11, 799–806. 26. Mans, R. J. and Novelli, G. D. (1961) Measurement of the incorporation of radioactive amino acids into protein by a filter-paper disk method. Arch. Biochem. Biophys. 94, 48–53. 27. Lee, K., Mossine, V., and Ortwerth, B. J. (1998) The relative ability of glucose and ascorbate to glycate and crosslink lens proteins in vivo. Exp. Eye Res. 67, 95–104.

Amino Acid Analysis Protocols (Methods in Molecular Biology)

Amino Acid Analysis Protocols (Methods in Molecular Biology Vol 159)

Amino Acid Analysis Protocols (Methods in Molecular Biology Vol 159)